Abstract

Summary

This study reveals that natural light color temperature cycles are irregular for over 90% of the year due to weather and atmospheric conditions, making current clear-day-only reproductions inadequate for biologically effective lighting. A TadGAN (generative adversarial network) model is proposed to generate realistic, irregular color temperature cycles, enabling more accurate dynamic lighting systems for human health applications.
Abstract

Key Findings

  • Irregular color temperature cycles were observed for more than 90% of the year due to regional weather and atmospheric conditions; regular cycles occurred only on some clear days.
  • The TadGAN-based generator was able to replicate the periodic characteristics of irregular natural light color temperature distributions, outperforming approaches limited to clear-day data.
  • Current dynamic lighting systems rely only on clear-day color temperature profiles, which the authors argue is insufficient to reproduce the actual beneficial effects of natural light on occupants.
Categories

Categories

The Science of Light: Proposes a TadGAN-based method to generate daily color temperature cycles that mimic irregular natural light patterns, contributing to lighting standards and dynamic lighting design.
Sleep & Circadian Health: Addresses the reproduction of natural light color temperature cycles, which are noted to significantly affect occupant health and circadian entrainment.
Authors

Author(s)

ST Oh, DH Ga, JH Lim
Publication Date

Publication Year

2022
Citations

Number of Citations

1
View more publications