Natural experiments from Earth Hour reveal urban night sky being drastically lit up by few decorative buildings
Chu Wing So, Chun Shing Jason Pun, Shengjie Liu, Sze Leung Cheung, Ho Keung Kenneith Hui, Kelly Blumenthal, Constance Elaine Walker

TL;DR
This study uses Earth Hour as a natural experiment to quantify the impact of specific decorative lighting sources on urban night sky brightness in Hong Kong, revealing significant reductions during targeted lights-out instances.
Contribution
It identifies key sources of urban light pollution and demonstrates the effectiveness of targeted lights-out measures using multimodal observational data.
Findings
Up to 50% reduction in night sky brightness during Earth Hour.
Major contributions from LED billboards and facade floodlights.
Weekend lighting patterns remain unaffected during Earth Hour.
Abstract
Light pollution, a typically underrecognized environmental issue, has gained attention in recent years. While controlling light pollution requires sustained efforts, Earth Hour offers a unique natural experimental setting to assess temporary lights-out measures. Using photometric and spectroscopic sensors, we observed up to 50% night sky darkening during Earth Hour from 2011 to 2024 in Hong Kong, primarily as a result of a small but critical number of lights-out instances in central business districts, as evidenced by crowd-sourced photography records. Weekend lighting pattern in the city during Earth Hour remained unaffected. The emission reductions mostly occurred in the 445-500, 500-540, and 615-650 nm spectral ranges-corresponding to peak emissions from LED billboard screens- and in the 585-595 nm range, associated with metal halide floodlights used for facades and billboards. Our…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
