SpaceX and other companies are still struggling to make their satellites darker in the night sky.
Nature | May 26, 2022
It’s been three years since SpaceX, an aerospace company in Hawthorne, California, launched its first batch of Starlink Internet-communication satellites, sparking concern among astronomers about the streaks the satellites leave in photographs of the night sky. Since then, many other Starlinks have launched: more than 2,300 of them now orbit Earth, comprising nearly half of all operational satellites.
Scientists have made some progress in coping with the onslaught. For instance, within days the International Astronomical Union (IAU) will debut a website including tools to help telescope operators predict satellite locations so that they can point their instruments elsewhere1.
But accumulating evidence reveals just how much these satellite ‘megaconstellations’ will interfere with astronomical observatories and other skywatchers around the globe. And satellite companies have not yet found a solution. SpaceX had been trying to address the problem by putting sun-blocking shades on its Starlinks to dim their appearance in the night sky. But Nature has learnt that the company has stopped doing so.
Tens of thousands of new satellites could be launched in the next few years. “This is an unsustainable trajectory,” says Meredith Rawls, an astronomer at the University of Washington in Seattle. “At the moment, our science is fine. But at what point will we miss a discovery?”
The toll of megaconstellations
Since the first Starlinks launched, astronomers have gone from panicking about the satellites photobombing scientific observations to organizing a global response. After a series of international workshops in 2020 and 2021, the IAU set up a Centre for the Protection of the Dark and Quiet Sky from Satellite Constellation Interference. Its soon-to-be-launched website is meant to serve as a hub for astronomers, policymakers, satellite operators and the public to coordinate on how to reduce the impacts of satellites blazing through the skies.
A recent study suggests that future satellite constellations will be the most visible during summer nights at latitudes of around 50 degrees south and 50 degrees north, where many European and Canadian astronomical facilities are based2. If SpaceX and other companies launch the 65,000 satellites they have proposed, bright dots will buzz across the skies all night long at those latitudes around the summer solstice, the study says. In the hours around sunrise and sunset, about one in every 14 stars visible to the naked eye will actually be a satellite.
“It’s really quite horrifying,” says Samantha Lawler, an astronomer at the University of Regina in Canada, who led the work.
Astronomical observatories that study wide expanses of the sky rather than focusing on individual celestial objects will be most affected. The Zwicky Transient Facility (ZTF), which surveys wide swathes of the sky using a 1.2-metre telescope on Palomar Mountain, California, had satellite streaks in 18% of its images taken during twilight in August 20213. And that number has gone up as satellite numbers have increased, says lead author Przemek Mróz, an astronomer at the University of Warsaw. He ran a preliminary analysis of ZTF data from April 2022 and found that satellite streaks affected about 20–25% of twilight images.
So far, the ZTF hasn’t had many of its measurements damaged by satellite streaks, in part because its image-processing methods can detect and mask satellite trails, Mróz says. But other observatories face larger challenges — particularly the Vera C. Rubin Observatory, an 8.4-metre-wide telescope funded by the United States and under construction in Chile. Because it will photograph the entire visible sky every three days, it will be particularly vulnerable to satellite streaks trailing through its images. Rawls and other astronomers are working on ways to mitigate the damage, such as algorithms that can identify and erase satellite streaks from the data. But fixing the data still takes a lot of time and energy. “It’s certainly eating my career,” Rawls says.
A busy sky
The growing number of satellites also threatens to negatively affect the field of radio astronomy and to increase the amount of space debris. Other, broader impacts could affect life worldwide: the presence of satellites contributes to a background glow in the sky that can disorient animals that rely on celestial navigation. Satellite streaks can also interfere with human knowledge systems, such as Indigenous knowledge systems that rely on information from dark skies to mark important events throughout the year4.
The growing threat of satellite constellations adds to other degradations of the night sky such as light pollution, says Karlie Noon, a PhD candidate in astronomy and an Indigeneous research associate at Australian National University in Canberra. “In the same way that our lands were colonized, our skies are now being colonized,” she says. “And this isn’t just Indigenous people.” She points out that companies have launched satellites without necessarily consulting the scientific community.
Some satellite operators have been working to mitigate the problem. Companies including SpaceX, OneWeb in London and Amazon’s Project Kuiper in Seattle, Washington, have met regularly with the IAU and national astronomical societies about ways to reduce the satellites’ impact. SpaceX has tested methods of dimming its Starlinks, including installing sunshades. The sunshades do reduce the satellites’ brightness5, but they appear to have been left off the latest generation of Starlinks. Those satellites, launched since September, use lasers instead of radio to communicate with one another, and the sunshades interfere with those communications.
SpaceX is instead working on other mitigations such as adding stickers or other materials to satellite mirrors to reflect light away from Earth, David Goldstein, an engineer with the company, said during a webinar hosted by the UK-based Federation of Astronomical Societies (FAS) earlier this month.
How well that might work is still being sorted out. An unpublished analysis of 102 observations of the brightness of Starlinks over time suggests that those from the new generation seem brighter than those known to have sunshades. However, they are not as bright as the original Starlinks without sunshades, says Anthony Mallama, a retired astronomer in Bowie, Maryland, who ran the analysis.
Meanwhile, OneWeb has launched 428 of a planned initial set of 648 satellites. They orbit at much higher altitudes than the Starlinks do — 1,200 kilometres compared with 550 kilometres. The satellites are typically fainter than Starlinks simply because they are farther away, but they can vary quite a bit in brightness depending on how they happen to be catching and reflecting sunlight.
One preliminary study of 50 OneWeb satellites during 2021 found that nearly half of them were a little brighter than the ‘safe’ limit specified by astronomers, says Jeremy Tregloan-Reed, an astronomer at the University of Atacama in Copiapó, Chile. OneWeb says it is committed to reducing the visibility of its satellites; it uses a telescope in Sicily to measure their brightnesses and is drawing on that information to design future satellites that are fainter, Maurizio Vanotti, OneWeb’s vice-president of space infrastructure development and partnerships, told the FAS webinar.
Amazon’s Project Kuiper, which would add more than 3,200 satellites, plans to launch its first 2 prototype satellites by the end of this year. One of them will contain a sunshade so that the company can compare its ability to dim the satellites’ brightness.
There are no laws regulating how bright satellites should appear in the night sky, although the IAU and other astronomical organizations have been pushing the United Nations to recognize the problem. Representatives from many nations will discuss protecting the skies at a meeting of the UN’s Committee on the Peaceful Uses of Outer Space that begins in Vienna on 1 June.
- Osborn, J., Blacketer, L., Townson, M. J. & Farley, O. J. D. Mon. Not. R. Astron. Soc. 509, 1848–1853 (2022).Article Google Scholar
- Lawler, S. M., Boley, A. C. & Rein, H. Astron. J. 163, 21 (2022).Article Google Scholar
- Mróz, P. et al. Astrophys. J. 924, L30 (2022).Article Google Scholar
- Lawrence, A. et al. Nature Astron. 6, 428–435 (2022).Article Google Scholar
- Mallama, A. Preprint at https://arxiv.org/abs/2111.09735 (2021).