Numerical investigation of internal solitary waves from the Luzon Strait: Generation process, mechanism and three-dimensional effects
C. Guo,X. Chen,V. Vlasenko,N. Stashchuk
DOI: https://doi.org/10.1016/j.ocemod.2011.03.002
IF: 3.293
2011-01-01
Ocean Modelling
Abstract:A fully nonlinear, non-hydrostatic model, MITgcm, is used to investigate internal solitary waves (ISWs) from the Luzon Strait (LS). As the ISWs in the South China Sea (SCS) have drawn more and more attention in recent years, they are studied in various ways, i.e., via remote sensing images, in situ measurements, and numerical simulations. The inspiration of this paper derived from the potential flaws of different numerical models that were employed to examine ISWs. In this study, we performed three-dimensional (3D) experiments with realistic topography and stratification, as well as with fully non-hydrostatic terms in the model, which was rather important for investigating the ISWs. Modeling results showed that baroclinic tides in the LS were essentially three-dimensional (3D), and that wave structures around two ridges in the strait were complicated with interesting internal oceanic phenomena. Several zonal cross-sections were chosen to illustrate vertical structures of zonal velocity field, and to show their meridional variances together with surface horizontal velocity gradients in order to highlight the advantages of 3D modeling with fully nonlinear, non-hydrostatic terms. Following Vlasenko et al. (2005) , analysis of two parameters (Froude number and slope parameter that is defined as the ratio of inclination of topography to slope of radiated rays) that govern generation regime indicated that internal waves produced in the LS were subject to a mixed lee wave regime rather than baroclinic tide regime or unsteady lee wave regime. The propagation of ISWs beyond the generation area showed that manifestation of 3D effects was not very obvious, which, through further analysis, was mainly attributed to homogeneity of topography, inaccuracy of barotropic forcing, and Kuroshio intrusion in the LS. To better understand the necessity of 3D modeling, we chose several zonal cross-sections and performed various sensitivity experiments to show discrepancies between 2D and 3D cases.