The importance of wave-induced mixing parameterizations was evaluated with a regional ocean model around Japan. Numerical experiments were conducted from March to September 2018, corresponding to the stratified period. The turbulence kinetic energy budget equation in control experiment (CR) adopted the Noh and Kim scheme, while the experiments LC additionally included the wave-induced mixing based on Langmuir circulation. The results were compared with Himawari-8 sea surface temperature data and GTSPP in situ observations of temperature and salinity profiles. The results showed that LC can effectively enhance the upper layer vertical mixing. By comparison, the experiment OM with wave-induced mixing based on wave orbital motion significantly overestimated the vertical mixing, which results in pronounced cold bias especially in the northern part of the model domain in summer. The comparison of mixed layer depth showed that the difference between LC and OM was relatively small for large wave motion, while only LC can realistically represent the vertical mixing for small wave case under various wind conditions. This is because the OM scheme accounts only for wave effects, whereas the LC scheme represents the combined influence of waves and wind. Furthermore, this study showed that the wave orbital velocity (WOV) and Langmuir number (La) had little correlation. It was implied that WOV alone cannot adequately represent the relative contributions of waves and wind to upper-ocean mixing, whereas La can serve as a more appropriate metric for estimating non-breaking wave-induced mixing.