Spectral energy, luminance and chromaticity of sky light in Rayleigh Atmosphere (in which the molecular scattering only is considered and no absorption) are varied by a lot of factors.
The light passing through Rayleigh Atmosphere is scattered in order by the air molecules and transmits in all directions according to Rayleigh's phase function.
In this paper, spectral energy of each scattering light up to the fourth order is estimated in the visible region (the wave length range from 0.38 to 0.75μm) for reflected light from the earth's surface, which is one of the principal factors of variation.
In the visible region, the more the ordinal number of scattering increases, the more the intensity of scattering light is weakened and the logarithms of the spectral energy of the scond, the third and the fourth order except the first order decrease nearly inversely to the increase of the ordinal number of scattering.
The error of the spectral energy due to neglect of the scattering light of a higher order than the fourth is estimated to be about 2.6% at the wave length 0.38μm and about 0.02% at 0.75μm.
The reflected light from the earth's surfase increases the luminance of the sky in the horizontal direction more than that in the zenith direction, and decreases the colour temperature of the former as compared with the latter.

View full abstract

The method developed in the authors' previouspaper is applied to assess ment various artificial daylights in the ultraviolet spectral range. Assessed results are compared with those of thecolor renderting property use of fluorescent samples foundin terstiege's and with those by the degree of fluorescence-excitation by Berger and Strocka. The assessment of color rendering property is significantly influenced by the difference in spectral power distribution between the test and the reference light source in the visible spectral range and is rejected as inadequate index for assessing the ultraviolet range of spectral power distributions of the artificial daylight. There is no correlation between two residual methods of assessment; the one by the degree of fluorescence-excitation and the other by use of fluorescent metamers proposed by the present authors. By inspecting the ultraviolet range of spectral power distributions of the tested light sources, it is discussed what can be detected by each of the methods. It is pointed out in the discussion that only onl, i, e, the method by the degree of fluorescence-excitation or the method by use of fluorescent metamers is insufficient and a combined treatment of both the methods is recommended for assessing the ultraviolet spectral range of artificial daylights.

View full abstract