When astronomers first dreamt up the Vera Rubin Observatory in the 1990s, the sky above the Chilean Cerro Pachón, where the star-observing machine was to be located, looked different than it does today.
Dotted with millions of stars, galaxies and nebulas, it was only occasionally crossed by a lone satellite. Then, just a few years before the observatory’s expected inauguration, the era of megaconstellations took off, and astronomers found themselves racing to find ways to protect the telescope’s images from satellite contamination. They didn’t have much time.
When construction of the $680 million observatory began in 2015, everything was still going according to plan. Four years later, SpaceX launched the first batch of Starlink internet satellites, Starlink trains became a thing, and astronomers realized that the satellites, orbiting only 340 miles (550 kilometers) above Earth, were too bright not to interfere with their observations. Vera Rubin, due to its wide field of view and exceptional sensitivity, was to feel their presence especially keenly.
“All of the characteristics that make Vera Rubin Observatory so amazing for surveying the whole southern sky also mean it’s going to see a whole bunch of these satellites,” Meredith Rawls, a research scientist for the Vera C. Rubin Observatory Legacy Survey of Space and Time (LSST) and an astronomer at the University of Washington, told Space.com.
The Vera Rubin telescope, which is set to open its eyes to the sky later this month, has a 26-foot-wide (8.4 meters) mirror, one of the largest in the world. It’s also fitted with the largest camera ever built for an astronomical telescope — a 5.4-foot-wide (1.65 m), 3,200-megapixel device with 189 CCD detectors.
The telescope will be able to tilt its mirror to change its view of the sky every night in order to complete a whole sky survey every three days. In each sweep of the celestial dome, the telescope will not only capture the myriads of stars and distant galaxies invisible to the human eye, but also the thousands of satellites that are millions of times brighter than those precious astronomical targets. And this problem will only get worse with time.
SpaceX was initially talking about a constellation of 12,000 satellites but now plans a fleet of 42,000 spacecraft. Other broadband operations — like Amazon’s Kuiper network and the Chinese projects Guowang, Qianfan, and Geespace — intend to launch tens of thousands of satellites of their own. Currently, about 10,000 satellites orbit Earth, but the number could increase to 100,000 in the next decade if all those plans come to fruition.
“It’s unfortunate that this huge increase [in the number of satellites] is coinciding with the decade of Vera Rubin’s operation,” said Rawls. “It’s existentially frustrating that we are putting a bunch of stuff in orbit that is interfering with our views of the cosmos.”
Rawls has worked on the Vera Rubin Observatory project since 2016, initially developing image processing algorithms to filter out faults of the camera’s sensors and detecting unexpected celestial phenomena such as supernova explosions. The arrival of Starlink and its counterparts forced her to refocus. Today, she develops techniques to flag the presence of satellites in images and distinguish them from objects of astronomical interest, including passing asteroids.
Still, Rawls said that the satellite streak problem is not a death threat for Vera Rubin’s science mission. She describes the satellite streaks more like “bugs on a windshield” on a summer night, obscuring the view at times, but not completely ruining it.
“It’s true that a large fraction of exposures is going to contain a satellite streak, but the field of view is big, and so the number of actual pixels that are affected is very small,” said Rawls. “At most, [the satellite streaks] are a few hundred pixels wide. But a single detector has 4,000 pixels, and the camera has 189 CCD detectors tracking the sky.”
Noelia Noël, a professor of astrophysics at the University of Surrey in the U.K., told Space.com that up to 40% of the images captured by the Vera Rubin telescope over its 10-year mission are expected to have streaks in them.
“If you take 10 million images, over 4 million of them could be degraded,” said Noël, who is also part of Vera Rubin’s LSST project. “This is a huge waste of taxpayers’ money. One night of Vera Rubin’s observations costs something like £60,000 [about $81,000]. So, if you ruin the images, it’s your money going to waste.”
Apart from outshining legitimate objects of interest, the satellites could also be mistaken for real celestial phenomena. In 2021, for example, a group of scientists thought that a star exploded in the oldest known galaxy when they observed a sudden brightening in images taken by the Keck Telescope in Hawaii. It later turned out that, as the astronomers pointed their instrument at the galaxy, a piece of debris passed in front of their field of few, reflecting sunlight.
“We don’t want to give people a catalog of data where each pixel is supposed to be an actual star, and then surprise, a third of them are just bright detections where it happened to be in the satellite trail,” said Rawls.
The algorithms developed by Rawls and her colleagues will use a stacking method to compare multiple images of the same portion of the sky to spot outliers and flag them. If a bright object appears in one image and disappears in the next, it’s more likely a passing satellite than a stellar explosion or dimming, said Rawls.
Megaconstellations like Starlink are only one part of the problem. In 2022, the American company AST SpaceMobile began deploying a constellation of its BlueBird satellites — essentially giant antenna arrays, each one covering 693 square feet (64 square meters). The satellites are intended to provide 5G via satellite directly to smartphone users on Earth, but they are also insanely bright. They are so bright, in fact, that the Vera Rubin Telescope must plan for their passes in advance in order to avoid them, according to Rawls.
“It would be a waste of 30 seconds looking at that portion of the sky with that super bright thing going through,” said Rawls. “Thankfully, there are not that many of these super big, super bright satellites yet. But I worry that might change in the coming years.”
How much of Vera Rubin’s precious sky views will be obscured by passing satellites and how much science will be lost as a result remains to be seen. Rawls hopes that attempts to darken satellites, already trialled by SpaceX with limited effects, will eventually succeed, reducing the light contamination to a minimum.
The International Astronomical Union (IAU) has previously called on satellite makers to strive to make their satellites invisible to the naked eye — an equivalent of magnitude 7 on the scale used to measure the brightness of celestial objects. The magnitude scale is inverse to the actual brightness and logarithmic, meaning that each subsequent grade is 2.5 times dimmer than the previous one. So far, Starlink satellites score between magnitudes 3 and 5.
“If satellite operators were able to keep their hardware within approximately the IAU brightness limit, then the impact on ground-based astronomy would be minimal,” said Rawls. “In practice, that’s not happening, because it’s really hard to make stuff that dark.”
Some glimpses of hope may be appearing on the horizon, however. U.K.-based company Surrey NanoSystems has recently introduced a new type of space paint that is easy to apply, resistant against the harsh space environment and reflects so little light that it could reach the needed brightness reduction. It may be ready just in time.
