Data-driven spectral sky models: A review

Abstract

Spectral power distribution of light plays a significant role in people’s comfort, visual performance, as well as their psychological and physiological health. Yet, most daylight simulations rely only on photometric quantities. This paper presents a differentiated colorimetric approach based on direction specific spectral data. It opens with a review of existing spectral sky models and a comparison study of their accuracy based on a large dataset of spatially, spectrally and temporally resolved measurements conducted at the TU Berlin. To conclude with the next steps towards creating a novel model to describe temporal variability and spatial distribution of spectral daylight characteristics.