Blurring the boundaries between Standard General Sky types due to multiple scattering of light

Abstract

The diffuse light reaching the ground varies with instantaneous physical state of the atmosphere, specifically with atmospheric turbidity, vertical stratification of attenuating and scattering constituents, cloud coverage and many others. Despite this complexity, a set of empirical models based on so-called homogeneous sky types is widely used to simulate the real lighting conditions. However, the sky type classification can be influenced by multiple scattering of light. Even if multiple scattering is essentially important under low elevation angles and high turbidity conditions, it is quite often ignored – e.g. when modelling the luminance patterns. As a consequence, the interpretation of experimental data can suffer from the omission of multiple scattering effects. Since multiple scattering smoothes the luminance patterns, the neighbouring sky types may be subject to blurring, and thus it might be difficult to distinguish them in automated data processing. In this paper we introduce a model that lowers the uncertainties caused by multiple scattering, so the subsequent numerical routines will definitely profit from more accurate identification of sky types.