2:15 PM - 2:30 PM
[AOS17-03] Laboratory experiment on vertical structure of mixing by nonbreaking surface waves
Keywords:ocean surface mixing
Turbulent mixing in the ocean surface layer is mainly caused by convection, wind and waves. Many recent studies suggest the importance of mixing caused by the nonbreaking surface waves (NBSWs). Langmuir circulations, which are vortex pairs generated by the interaction between wind induced shear flow and surface waves (CL2 mechanism), are one example of the mixing by NBSWs. On the other hand, some laboratory (Babanin and Haus, 2009; Dai et al., 2010; Savelyev et al., 2012) and numerical (Tsai et al., 2015, 2017; Fujiwara et al., 2020) experiments showed that mixing is also induced only by waves without wind.
Savelyev et al. (2012), hereafter referred to as S12, visualized structures at the water surface using infrared light and identified streaks in the presence of NBSWs, implying the vortex pairs near the surface. Tsai et al. (2017) compared the wave numbers of the streaks observed in S12 with the unstable wave number of CL2 instability and suggested that these streaks are vortex pairs generated by the CL2 instability. Laboratory experiment of Dai et al. (2010), hereafter referred to as D10, observed that initially stratified water was destratified under the NBSWs for several tens of minutes. We conducted direct numerical experiments of D10’s laboratory experiment (Imamura et al., 2021; Imamura and Yoshikawa, 2021) and confirmed that NBSWs produced vortex pairs generated by the CL2 mechanism and induced vertical mixing without wind. However, the deepening speed of mixed layer was slower in our numerical experiment than in D10’s laboratory experiment. The purpose of this study is therefore to confirm whether the CL2 mechanism really generates vortex pairs and induces mixing in the laboratory experiment.
The laboratory experiments were conducted at the wave tank (65m long, 5m wide, and 7m deep) of the RIAM, Kyushu University. The tank is inside a building so windless condition can be realized. A wave with a wavelength of 156 cm and an amplitude of 4 cm was generated.
First, to confirm the mixing by NBSWs, water temperature was measured from the surface to 90 cm depth before and at 4, 19, 34, and 49 minutes after the onset of waves. At 4 minutes after the onset, the temperature from the surface to a depth of 20 cm was destratified. On the other hand, at 19, 34, and 49 minutes after the onset, the temperature decreased at all depths. Such evolution was qualitatively similar to that of D10. While the decrease in temperature after a long-time wave propagation was observed both in the D10 and this laboratory experiment, it was not observed in our previous numerical experiments. This suggests that such a temperature decrease is not due to the mixing by NBSWs but due to the secondary circulation in the tank.
Next, we used microbubbles as tracer to see the structure of water motion. The vertical motion of the water was observed from 2.3 s to 21.0 s after the passage of the first wave. It was observed that the tracer formed a convex structure below the surface and went downward locally. The width of the structure became wider with time. Here, as Tsai et al. (2017), the wave numbers of the tracer structure were compared with the unstable wave number of the CL2 instability. The wave numbers of the structure evaluated to 5s after the first wave passed are consistent with the unstable wavenumber of CL2 instability.
These results show that the water temperature evolution in this laboratory experiment is qualitatively similar to that in D10’s. Vortex pairs were observed in our laboratory experiment, and they were likely generated by the CL2 mechanism. The time evolution of water temperature suggest that mixing is caused by the NBSWs as soon as the waves were generated (4 minutes later), while after a longer time, it may be caused by the factors other than the NBSWs such as secondary circulation in the tank. The overestimation of mixing in the laboratory experiment may be due to this secondary circulation and other factors.
Savelyev et al. (2012), hereafter referred to as S12, visualized structures at the water surface using infrared light and identified streaks in the presence of NBSWs, implying the vortex pairs near the surface. Tsai et al. (2017) compared the wave numbers of the streaks observed in S12 with the unstable wave number of CL2 instability and suggested that these streaks are vortex pairs generated by the CL2 instability. Laboratory experiment of Dai et al. (2010), hereafter referred to as D10, observed that initially stratified water was destratified under the NBSWs for several tens of minutes. We conducted direct numerical experiments of D10’s laboratory experiment (Imamura et al., 2021; Imamura and Yoshikawa, 2021) and confirmed that NBSWs produced vortex pairs generated by the CL2 mechanism and induced vertical mixing without wind. However, the deepening speed of mixed layer was slower in our numerical experiment than in D10’s laboratory experiment. The purpose of this study is therefore to confirm whether the CL2 mechanism really generates vortex pairs and induces mixing in the laboratory experiment.
The laboratory experiments were conducted at the wave tank (65m long, 5m wide, and 7m deep) of the RIAM, Kyushu University. The tank is inside a building so windless condition can be realized. A wave with a wavelength of 156 cm and an amplitude of 4 cm was generated.
First, to confirm the mixing by NBSWs, water temperature was measured from the surface to 90 cm depth before and at 4, 19, 34, and 49 minutes after the onset of waves. At 4 minutes after the onset, the temperature from the surface to a depth of 20 cm was destratified. On the other hand, at 19, 34, and 49 minutes after the onset, the temperature decreased at all depths. Such evolution was qualitatively similar to that of D10. While the decrease in temperature after a long-time wave propagation was observed both in the D10 and this laboratory experiment, it was not observed in our previous numerical experiments. This suggests that such a temperature decrease is not due to the mixing by NBSWs but due to the secondary circulation in the tank.
Next, we used microbubbles as tracer to see the structure of water motion. The vertical motion of the water was observed from 2.3 s to 21.0 s after the passage of the first wave. It was observed that the tracer formed a convex structure below the surface and went downward locally. The width of the structure became wider with time. Here, as Tsai et al. (2017), the wave numbers of the tracer structure were compared with the unstable wave number of the CL2 instability. The wave numbers of the structure evaluated to 5s after the first wave passed are consistent with the unstable wavenumber of CL2 instability.
These results show that the water temperature evolution in this laboratory experiment is qualitatively similar to that in D10’s. Vortex pairs were observed in our laboratory experiment, and they were likely generated by the CL2 mechanism. The time evolution of water temperature suggest that mixing is caused by the NBSWs as soon as the waves were generated (4 minutes later), while after a longer time, it may be caused by the factors other than the NBSWs such as secondary circulation in the tank. The overestimation of mixing in the laboratory experiment may be due to this secondary circulation and other factors.