A Fuess-Robitzsch-actinograph No.13123 and a Hasegawa-actinograph No.2436 of the same type were calibrated, by comparing with an Ångströmpyrheliometer No.234 and a silver-disc pyrheliometer No.754 which was designed by Prof. R. Sekiguti and made in the C. M. O. workshop. The observations were made during the period from March, 1937, to March, 1938.
According to the difference of the spectral distribution between the sun- and the sky-light, a correction should be added to the constant of the actinographs of this type, which is determined by the comparison with the pyrheliometer. As Dr. W. Mörikofer and others showed, the constant of this actinograph given by Prof. M. Robitzsch was proved to be too large. The correcting factor obtained by the direct comparison with the pyrheliometer only was 0.80 in the mean, and it becomes about 0.09 smaller, when we take the correction ξ/η into consideration. The latter correction was not exactly determined, but estimated by comparing the intensities of sky-radiation observed by our actinographs with the results obtained by Drs. A. F. Moore and C. H. Abbot on Mt. Hump in 1917. The intensities of sky-radiation computed from readings of Fuess No.13123 were, in any condition of the atmosphere, larger than the results obtained by Hasegawa No.2436. This difference might be explained by the difference of the thickness of glass-globes and of the reflective power of radiation in the range of long wavelengths. Fuess No.13123 has a glass-globe of 1mm thickness, and surfaces of its bimetals have been somewhat spoiled by the long use. On the other hand, the thickness of glass-globe of Hasegawa No.2436 is 3mm, and the sensitivity of this actinograph is larger. The results of our observation of sky-radiation agreed with those of Moore and Abbot's observation in better approximation in Hasegawa No.2436 than in Fuess No.13123. We, therefore, assumed, as the first approxiamtion, that in Hasegawa No 2436 ξ/η was equal to unity and the results obtained with this actinograph were more trustworthy. This negative correction is larger than the results given by H. Stapf in Met. Zeit, 1938. The future problem is to study the correction ξ/η more fully.
The variations of the constant with the altitude and the azimuth of the sun were studied also. It is found, besides the same results obtained hitherto, that the constant increases as cos
z becomes smaller than 0.44. The causes of these variations were also studied in detail. The greater value of the constant in the higher altitude of the sun could not be explained on the theoretical ground. But it is likely that the cause may be sought in the cooling effects due to the conduction of heat, convection currents, long-wave radiation, etc.
The constant must be changed according to the condition of the atmosphere, and under the entirely overcast sky it seems preferable to use the value obtained directly from the comparison with the pyrheliometer. It was also found that the effect of the radiation incoming through the glass-window is not large.
View full abstract