Study On The Influence Of Different Vertical Mixing Parameterization Schemes On Internal Wave Simulation Under Non-Hydrostatic Conditions

Internal waves are a ubiquitous dynamic process in the ocean,which has a significant impact on ocean mixing,marine ecological environment,marine engineering and military affairs.At the same time,with the development of ocean models,non-hydrostatic numerical models have become one of the common methods to study internal waves.Considering that the sub-grid physical process,such as turbulence,cannot be directly simulated,the model needs to use parameterization schemes to introduce its physical effects to improve the accuracy of model.In order to explore the influence of different vertical turbulence mixing parameterization schemes on internal wave simulation under non-hydrostatic conditions,this study firstly selects a non-hydrostatic ocean model independently developed in China — the MERF V2.1(Marine Environment Research and Forecasting Model,V2.1).Then,the model is coupled with two turbulence mixing parameterization schemes which are commonly used in the world — the MY2.5scheme(Mellor and Yanmada)and the KPP scheme(K-profile parameterization).Thus,the MERF has been developed into version 2.2 with the mature turbulence mixing parameterization scheme.Based on MERF V2.2,the internal lee wave on the supercritical Gaussian terrain is numerically simulated under the non-hydrostatic and hydrostatic approximation conditions,respectively.The distribution of temperature,horizontal velocity,vertical velocity and dynamic pressure that characterize the internal wave structure are compared and analyzed from different time scales(special time and period average).The results show that under non-hydrostatic conditions,both MY2.5 and KPP schemes can present part of internal lee wave signals at specific moments.On the whole,except for the temperature,the characterization of the other three physical quantities by the KPP scheme is more reasonable qualitatively and quantitatively than the MY2.5 scheme.Compared with the non-hydrostatic results,the internal wave amplitude simulated by the MY2.5 scheme under hydrostatic conditions is smaller and the wavelength is longer;however,the wavelength simulated by the KPP scheme is almost reduced by half.The more intense the internal wave activity,the more obvious the simulation difference between the hydrostatic and non-hydrostatic models.For the simulation of small-scale internal waves,the influence of the vertical turbulence mixing scheme is more significant than the choice of hydrostatic/non-hydrostatic model.In addition,based on the internal wave energy and turbulence mixing analysis methods,the effects of the above two vertical mixing schemes on the internal wave energy propagation and the turbulence vertical mixing at different time and space scales under the non-hydrostatic conditions are studied.From the perspective of the whole system,in the baroclinic energy(BCE),the baroclinic kinetic energy(BKE)of MY2.5 and KPP schemes are 55% and 58% higher than the available potential energy(APE)respectively,which is lower than the simulation result of constant mixing coefficients(72%).When the boundary layer is removed from the KPP scheme,the BKE accounts for nearly 110%.From the perspective of horizontal position,the ratios of BCE to barotropic energy(BAE)of MY2.5 and KPP schemes range from 5% to30%.The two schemes have little influence on the APE,mainly affecting BKE near the steep slopes,at the same time producing deviations on the prediction of internal wave generation range.The mixing caused by the MY2.5 scheme mainly occurs in the region of internal wave generation(sill steep slope),which weakens the simulated internal wave intensity;the mixing caused by the KPP scheme mainly occurs at the top of the sill,that is,the boundary layer effect leads to intense mixing in the shallow water area.