摘要
Due to the difficulty and high cost of direct observation of turbulence mixing, the study of turbulence mixing is largely limited. Therefore, the method of estimating ocean turbulence mixing based on temperature, salinity, and current data came into being. In this paper, for the first time, we evaluated the applicability of currently used Gregg-Henyey-Polzin (GHP) fine scale parameterization, Mackinnon and Gregg (MG) parameterization, and Thorpe scale methods using data of free-falling microstructural turbulence profilers from 23 observation sites and hydrological observation data in the Luzon strait. It was found that the GHP parameterization method could well estimate the turbulent mixing of Luzon Strait. Although the dissipation rate estimated by the GHP parameterization method is generally weaker than the observation results, the difference between the estimation and observation results within a factor of 2 accounts for 71%, showing the same distribution characteristics with the horizontal and vertical distribution of the dissipation rate observed by the microstructural turbulence profilers. Based on the MG parameterization method, it was found that the dissipation rate of shallow than 1200m on the west side of Luzon Strait was larger than the observed value, but the distribution characteristics were the same on the whole. In addition, the results of MG parametric estimation and observation were within a factor of 2, accounting for 58%. It shows that compared with the GHP parameterization method, the estimated value of MG parameterization method deviates more from the observed value. When the Thorpe scale method was used to estimate the dissipation rate of Luzon Strait, only 30% of the difference between estimation and observation was within a factor of 2, and 70% of the results were more than one order of magnitude, showing the spatial distribution is quite different from the observed results. Comparing the results of turbulent mixing parameterizations in Luzon Strait, showing that the GHP parameterization method is the best, the MG parameterization method is the second, and the Thorpe scale method is inferior to the GHP and MG parameterization methods.
-
单位中国科学院; 中国科学院大学