The last time NASA went to the moon, the Apollo 17 astronauts brought back the famous “Blue Marble” photograph of the Earth, as seen from 18,000 miles away. Fifty-three years later, it was time for an update. This week, the Artemis II crew beamed back its own version of the Blue Marble, titled “Hello World.” Though they used a Nikon D5 instead of a 70mm camera, the resulting picture was, in many ways, stunningly similar to the original. Each shows our garden planet, dizzyingly small and bright against the vast blackness of space, looking peaceful and untouched.
Were “Blue Marble” and “Hello World” taken from the night side of the Earth instead, the pictures would differ dramatically. In the 25 years between 1992 and 2017 alone, detectable artificial light on the planet has grown by an estimated 50%. Many parts of the world are expected to lose more than half their visible stars within a generation due to light pollution. Night is “being lost in many countries” due to the onward and upward march of progress and development, the BBC warned in 2017. Researchers have even found that plants, animals, and microbes release more planet-warming carbon dioxide under artificial nighttime light than their counterparts in the natural dark.
But our assumption that the planet is brightening gradually and unidirectionally is flawed, according to a new study in Nature, published today. While prior research on artificial light tracked its changes across months or years, University of Connecticut Professor Zhe Zhu and his colleagues analyzed nearly 1.2 million daily images captured between 2014 and 2022 by NASA’s Black Marble Night Time Light satellites to deepen their understanding of the day-to-day dynamics. “It’s an entirely different perspective” on artificial light compared to what existed before, Zhu told me. “I’d say I’m quite shocked.”
Rather than illustrating the well-known story about the Earth getting brighter, Zhu’s research describes a dynamic, volatile, flickering planet. Though the researchers confirmed a 34% overall increase in brightness during the study’s nine-year scope, it was offset by an 18% dimness, meaning the net increase in brightness was only 16%. Further, nearly half of the portions of land area that experienced at least one change in artificial light also experienced some form of abrupt change — that is, a brightening or dimming event that unfolded over weeks or months rather than years, such as grid failures in Venezuela, load-shedding in South Africa, changes to fossil fuel operations in Texas, and armed conflicts such as the war in Gaza. “We see all the ups and downs of human civilization reflected on a daily basis” in the artificial light, Zhu said.
The data also illustrates the planet’s energy story. In Western Europe, for example, Zhu expected to see increased brightness due to the region’s high level of development. “But it’s entirely the opposite,” he said. “All of Europe is in a dimming area.” The researchers looked more closely and attributed the change largely to the switch to LED bulbs, which are better at directing light at the ground. (Some argue that LED lights have created a kind of Jevons paradox by making affordable lighting cheaper and more widespread; a 2017 study led by Christopher Kyba from the GFZ German Research Centre for Geoscience also argued that LEDs are making skyglow worse because they scatter blue light. Zhu acknowledged that the satellites used in their study do not pick up blue light, which is part of why the transition shows up as a dimming effect; Kyba is a coauthor on the study.)
Dimming in Europe is also attributable to regional energy conservation-related policies. The dimming patterns mapped neatly onto national borders, such as a 33% net drop in France and a 22% drop in the U.K.
Dimming and brightening aren’t necessarily negative or positive indications, though. The Midwest, for example, has dimmed due to economic contraction, including declining urban cores and manufacturing sectors. But like Europe, the region has also implemented energy-efficient lighting programs. In places like West Africa and parts of Asia, brightening indicates improvements in energy access and investment in new infrastructure; in other parts of the world, it might indicate increased fossil fuel extraction, as evidenced by abrupt changes in brightness caused by flaring. Overall, though, artificial light has outpaced population growth. “Each person is emitting more light, so efficiency is, at the global scale, mostly decreasing,” Zhu said.
Beyond the broadest conclusions, using averages in nighttime light as a proxy for economic growth and energy use, as researchers have for decades, is grossly inaccurate, Zhu’s paper shows. His research will likely have significant implications for those working in fields that touch on animal migration, insect behavior, and human health as they relate to light pollution — fields where the assumption had been a gradual increase in light, rather than the more complicated picture of abrupt shifts and changes.
Zhu and his team were working on a three-year grant that has since expired. But he said he hopes that one day there will be a near-real-time, publicly available version of the Dark Marble satellite images his team used for their research. Such a tool would have obvious applications in the climate, energy, and humanitarian sectors, from monitoring energy access, natural disasters, and emissions to tracking regional policies — especially since existing averages, which only relay long-term trends, miss the more dynamic ebbs and flows of our energy infrastructure at the global scale. “It’s not just people using more lights,” he said of the findings. “It’s a tug-o-war.” You just have to know how to look.
