Wind velocity and temperature profiles and fluctuations in wind direction were observed over a rice field in Sept. 17-18, 1955. The mean wind velocity is measured with 10 small cup anemometers distributed logarithmically to a 6m. height pole, and thermistor thermometers are used for the measurement of temperature distribution.
Under the neutral thermal condition of the wind velocity of 1-5m/sec. at 2m. height, the zero-plane displacement
d decreased and the roughness parameter
z0 increased respectively with the increase in wind velocity. These facts are contrary to results reported before (Tani, Inoue and Imai 1954), the cause of this discrepancy seems to be due to the different condition of rice stalks in growth stage, density, flexibility etc.
A wind direction-meter with 50cm, long arrow made with aluminum plate sends the signal of direction angle to the recording system by means of an A. C. selsyn-motor. The recording chart is of 12cm. width which is corresponding to 130° direction angle, and the recording speed is regulated to be 2mm/sec. Recorded charts are read at every 2.5 seconds and these values are made use of in the following analyses.
The hystograms of wind direction show the normal distribution pattern with some excepts. The range of flucuation during 5min. is about 30° under the condition of small turbulence and 60-70° under that of large turbulence.
The lateral turbulent velocity ‹
v2›
1/2 is in proportion to the mean wind velocity U, but under the unstable condition it becomes larger than under the neutral or stable condition.
The lateral turbulent energy ‹φ
2› and the autocorrelation
R(t) with respect to the angle fluctuation φ are calculated for some 3 cases, and the following results are obtained:
(1) 1355-1400, 17th. The results indicate that the reading interval 2.5sec. of chart is too small to obtain the characteristic nature of the wind turbulence concerned.
(2) 1540-1545, 17th. The observation was carried out under the stable condition. The scale of the horizontal largest turbulon seems to be decreased so that the averaging time of 5min. seems almost comparable to the passage-time of that largest turbulon.
(3) 1343-1348, 18th. In this run, 1-R(t) shown to be in proportion to t
2/3 as expected theoretically, and the characteristic time
T0 at which R(t) becomes zero is shown to be about 60sec.
Considering that this characteristic time is much larger than the passage-time of the coupling turbulon directly influenced by the ground surface, which is estimated to be about 10sec., the authors suppose that the turbulons observed here are not the coupling ones but the topographical ones of the medium scale.
View full abstract