The height of the sea level changes as the atmospheric pressure varies. Their phases, however, do not always coincide, but generally the level change lags behind the pressure change by a few hours. Moreover, Dr. S. Ogura found that this time retardation seems to increase with latitude in the neighbouring seas of Japan. As the result of investigation of the velocity distribution with depth in case of tidal currents and high tides caused by surface pressure and of the decay of long waves sent out from a wave source, the present author showed that we may consider a resistance proportional to the velocity of fluid particle in place of eddy viscosity as a frictional force acting on a long wave. Using this particular form of resistance and assuming that the motion is simple harmonic, the equation to be solved is where ζ* is the equilibrium elevation, h is the depth of sea and k, λ are the coefficient of resistance and that of Coriolis' force respectively. By solving this equation, we have the phase difference ε between surface pressure and the corresponding change of sea level, that is approximately, where T and U are the period and velocity of surface pressure respectively. Since the result of calculation by the above formula essentially agrees with that of observations, we may conclude that the said phase retardation is due to the resistance acting on sea water. Moreover, we can explain the reason why the time retardation increases with latitude by considering the depth of sea, velocity and period of surface pressure and Coriolis' force. Though in the above discussion we have assumed a simple harmonic variation of surface pressure, it can be shown that the conclusions above obtained are also valid in case of a solitary wave of surface pressure by using Fourier's integrals. And the author showed a method to discuss the level change in case of an arbitrary type of change of surface pressure by using Fourier's transformation. In case of a finitely extended ocean, the phase retardation is larger or smaller by ten to twenty percent than in case of an infinitely extended ocean according to the relations between geographical features and the direction of propagation of surface pressure. Thus, it was found that the effect of land is small in our present problem.
The gustiness of wind is one of the most effective representative of the state of wind. But it depends on the method and instrument of observation. Generally, it is thought that the fluctuation of wind velocity is caused by the eddies in the wind. Then the gustiness of wind relates to the range of the observable eddies, which is limited by the duration of observation and the sensibility of instrument. The duration of observation regulates the effective largest eddy, and the sensibility of instrument regulates the effective smallest eddy. The negative five-thirds power law of eddy spectrum is assumed to be applicable to these eddies. Then the relation between the duration T and the gustiness is obtained as follows: These results are ascertained by the comparison with practical observation. The influence of the sensibility of instrument is also estimated.
The Pacific coast of Northern Honshû is known from historic times as the region most frequently attacked by destructive seismic sea waves or Tsunami. Even in recent years, i.e., in 1896, 1933, etc., this district was violently damaged, but the establishment to prevent damages is not yet complete. Every time after the destructive Tsunami the prevention from damage was considered and harbour constructions were made at various places in this district. It is not needless, therefore, to study the effect of these constructions upon Tsunami. As a first stage we made a model experiment of Tsunami which occurred in Shizukawa harbour. The ratios of the model to the actual bay in horizontal and vertical lengths were 1:1500 and 1:125, respectively. The model was put into a wooden tank 240cm long, 120cm wide and 60cm deep; and waves were generated by an iron plate by pulling or pushing it longitudinally. Waves thus generated and propagated in the bay were measured by small tide-gauges specially deviced for this experiment. The results of this experiment are as follows. (1) The height of the wave is highest at the head of the bay, being about 4 times the height at the mouth of it. By the aid of breakwaters built after 1933 the height of wave was lowered in the harbour and at the eastern side of it. However, at the western side where waves were heighest in 1896 and 1933 the wave seems to have become higher than before. (2) It takes about 4 minutes for the wave to reach the head of the bay from the mouth. (3) The secondary undulation of the bay was induced by the incident wave and its observed period was 11.8 minutes. On the other hand, the period deduced from the depth and the form of the bay is 13.3 minutes. (4) It is found that the height of the crest of the secondary undulation decreases exponentially with time. As an empirical formula we obtained h=h0e-0.039t, the unit of time being second. Comparing the formula with the result which was obtained by one of the present authors on the effect of viscosity on Tsunami, we evaluated the value of the turbulent viscosity of the sea water as 17 which is considered to be a reasonable one for the water in the bay. Finally as it is important to prevent the wave at the head of the bay to protect the centre of this town from Tsunami, we studied the method of prevention from damage at this place constructing some model breakwaters there. The details will be printed in Science Reports of the Tôhoku University, Series 5, Vol. 1, No.2.
The author has studied the relation between the wind velocity and the inclination of a cylindrical streamer made of cloth. The ordinary theory of aerofoils treats only when the angle of incidence is small. In the present problem the angle varies between 0 and 90 degrees, so that the theory cannot be applied to it. Therefore, the result of the theory of discontinuous current and the coefficient of interference of the two wings of a biplane are used to compute the pressure on a cylinder placed obliquely in a current. By this computation considering the pressure and weight of the streamer, the equation (c) of §2 showing the relation between the inclination and the wind velocity was derived. By comparing the results of this computation and of the actual observation, it may be concluded that this computation is reliable to some extent.