부산대학교 도서관

The observational evidence for atmospheric tides is considered in the first chapter. In the Taylor-Pekeris theory some terms involving the vertical velocity and the horizontal component of the coriolis force are ignored and in Section 1 of Chapter 3 one special case is con-sidered where these terms can be included. The solution shows that the corresponding Taylor-Pekeris solution is invalid. The second section of this chapter is con-cerned with the annual variation of the natural period near 12 hours and it is concluded that the semi-diurnal wave, S2 is not sharply resonant. An attempt is made to find the effects of the average zonal wind and meridional temperature gradient on the natural period near 12 hours, but it is found that their effect is not significant, on the theory presented. A significant correlation is obtained between the phase of S2 at Accra and the zonal index, but surprisingly no significant correlation with amplitude was discovered. The cause of the annual phase change of the semi-diurnal lunar tide L2 is considered. It is concluded that L2 does not seem to be influenced at ground level by the natural period near 12 lunar hours, and that the observed phase change must be produced primarily by an annual change in damping in the troposphere. There follows a qualitative discussion of the regions of heat input for tho 8 and 6 hour waves, depending on the observed phase difference between the first symmetric and antisymmetric latitudinal modes. The chapter is concluded with some considerations of the amplitude and phase of that part of S2 produced by tropospheric heating. The Holmberg hypothesis is introduced in the final chapter. It is shown that the 24, 12 and 8 hour waves cannot be highly resonant on the planet Mars. However it is pointed out that large magnifications are not necessarily required for the Holmberg mechanism to operate. It is not unlikely that the mechanism operates on Mars at present because of the small dissipation.