[AOS09-P21] Numerical studies on asymmetric responses of the upper ocean to tropical cyclone
Keywords:Tropical cyclone, Vertical turbulent mixing, Ekman suction, Rightward bias
The SST decrease during TC passage is mainly caused by two processes; vertical turbulent mixing and advection (Ekman suction). The rightward bias is associated with the mixing which is more intense on the right side of the TC track due to resonant inertial currents with winds. Magnitude and distribution of SST decrease induced by the mixing should vary with parameters characterizing TC; transition speed UT, radius of maximum wind Rmax, the friction velocity U*, and the Coriolis coefficient f. However, detailed physical processes of the mixing and its dependence on these parameters have not been well quantified, because of lack of in-situ data and complicated processes involved in the mixing.
The present study focuses on the vertical turbulent mixing effects on asymmetric SST response. To this aim, vertically one-dimensional model was used in which the vertical turbulent mixing was parameterized with the Nakanishi-Niino-Mellor-Yamada scheme modified by Kitamura (2010). We considered TC-like wind stress of magnitude U*2 at radius Rmax from the TC center moving with transition speed UT and investigated the mixing on both sides of TC track at distance r. Experiments with several friction velocities U*, maximum wind radii Rmax, transition speeds UT and distances r were performed. No surface heat flux was considered for simplicity.
The results showed that the rightward bias of the SST response depended on these parameters. The location of the largest SST decrease moved away from the axis as UT increased, but it depended less on U*. The SST decreased more with larger U* and smaller UT. These results can be explained by energy flux at the surface. In our model configuration, the energy flux depends on two non-dimensional parameters, UT/fRmax and U*/UT, and the energy flux had maximum and turbulent kinetic energy penetrated deeper at larger U*/UT and at UT/fRmax~1, suggesting the resonance between wind and the inertial current. The SST decrease corresponded with the energy flux, but it depended largely on U*/UT, implying importance of wind forcing duration.
Thus, the asymmetric SST response due to the vertical turbulent mixing can be characterized with UT/fRmax and U*/UT. Additional advection effects on the SST response will be discussed at the conference using three-dimensional simulation in which the advection as well as the mixing are fully included.