For the purpose of high water forecast, an automatic tide station which can send radio signals and can be operated at remote position, has been developed. The gauge consists of the transmitter and the receiving set. Both sets are controlled by the clocks, and the height of the tide is recorded on the chart once an hour. The frequency of the radio wave is 414 m_??_ sec-1. The principle of remote recording is that of so-called Olland system. Synchronization of the transmitter and the recorder is caused by a small permanent magnet motor with a speed regulater. The automatic zero adjust of the record is made by a long sign of the signal in every cycle, and the height of tide is marked with a dot by the short signs on the chart. These equipments have been tested at the Central Meteorological Observatory and the Tsukishima tide station in Tokyo, from Oct., 1953. The accuracy of the measurement of tide height was within about 1%. Power required for the transmitter is 6V 3W, for which 6 pieces of 500 AH air wet depolarized cells are used.
Under considerations of natural ventiration caused by the temperature difference between room air and open air, and total heat transmission through the room-walls, the author obtained theoretically the following equation, where θ and T denote temperature of room air and open air respectively, t time and β', η' constants. In this paper the author determined analytically in several cases the value of η' and β' which indicate the capacity of cold-protection of house, comparing the solution of the above equation with observational data of θ and T. The coincidence between the results obtained analytically and those obtained from the observation was considerably good.
The author computed the intensity of scattering at 5000m level in four wave-length ranges, and obtained the intensity distribution on the blue sky dome. Each of four partial solar radiation's energies contained in the wave-length ranges 0-0.5190 μ-0.7077 μ-1.0575 μ-∞ is equal to one fourth of the total energy and are denoted by I, II, III and IV. Let h be the sun's altitude, A and θ the azimuth relative to the sun and the altitude of the portion of sky dome, respectively. The intensity of first scattering increases with decreasing wave-length for each A and θ except for h=0 and θ=0, where the domain II has the greatest intensity from A=0 to A=120°, but III in A=150°, 180°. For the altitude higher than h=30°, the intensity decreases with increasing θ for each range and A, for h=0 in domain I it is greatest at θ=30° and decreases to both sides from it, further in domains II, III and IV it decreases with increasing θ. The conditions of secondary scattering is the same with the primary for h_??_30°. In h=0, the intensity decreases with increasing θ in each range and A, and II has maximum for all A and θ. The conditions of the total scattering are identical with the primary. The intensity of total scattering for total wave-length range decreases with increasing θ for all h. The horizontal intensity due to total scattering increases with increasing altitude and decreasing wave-length, and therefore the value for total wave-length range increases with increasing altitude. The horizontal intensity due to total radiation increases with increasing h for each range, and decreases with increasing wave-length for h=0, but it is inverted for h_??_30°.
We made a quantitative forecast of precipitation based on the numerical prediction method, using three layer model of the atmosphere. Treating the transport of water vapour three dimensionally and considering the temperature change, the excess of water vapour over saturation was estimated, which we assumed to fall off entirely as precipitation. These calculations were performed by the graphical method designed by Fjφrtoft, using the geostrophic approximation for the wind velocity. The North American Continent was chosen as the forecasting a_??_ea. For 12 hour forecast, the coincidence between the observed and forecast precipitation was found to be fairly good both in amount and area. This result suggests the possibility of forecasting weather based on the large scale motion of the air. Lastly, brief discussions were made on the non-adiabatic effect, which we disregarded in our practical forecasting.
During the period from August to September in 1954 the investigation of substances in raindrops and of aerosols in the air were made at several stations in the central part of Hokkaido. The results obtained are as follows. (1) The size distribution of the particles is expressed by a formula dN/dr=kr-4 for the range of particle radius r below 10μ, and for the particles of r>10μ we have a formula dN/dr=k'r-5. (2) The number of large hygroscopic nuclei in the air is about 500/m3. (3) The number of large non-hygroscopic particles of radii larger than 15μ in the air is about 4000/m3 and it seems that these non-hygroscopic particles also serve as the nuclei of raindrops. (4) The mass of non-hygroscopic particles in raindrops is considerably larger than that of hygroscopic particles. (5) The larger non-hygroscopic particles are captured by raindrops more easily than the smalle_??_ ones.