Starlink interference threatens radio astronomy’s golden age

In what researchers are calling the most comprehensive survey of its kind, scientists at Curtin University in Australia have cataloged more than 112,000 unintended radio emissions from Starlink satellites. These emissions now regularly disrupt astronomical observations and may jeopardize the future of radio astronomy.

New study reveals high level of Starlink interference

The study, conducted by the Curtin node of the International Centre for Radio Astronomy Research (ICRAR) and published in Astronomy & Astrophysics, expressed concern over a high level of radio interference from the satellite constellation Starlink. It analyzed 76 million images from a prototype station for the Square Kilometre Array (SKA), soon to be the world’s largest and most sensitive radio telescope. Over just four months, the team detected interference in up to 30 percent of their images, often in frequencies that are supposed to be protected for science.

Study co-author Steven Tingay, the executive director of the Curtin Institute of Radio Astronomy, explains the impact of the interference on the SKA to Astronomy, saying, “For some of the SKA’s science goals, I expect that if the current situation persists, the science will still be manageable. In some areas, the impacts may not be that great. In some other areas, where the science demands the most sensitive observations and contamination-free data, I personally think that the impacts may be significant.”

Starlink, operated by SpaceX, is the largest satellite internet constellation in orbit. Its thousands of spacecraft deliver broadband from space, but as this study shows, they also contribute to a growing source of radio interference. The problem isn’t just Starlink’s size – though the network had over 7,000 satellites in orbit at the time of the study, with 477 new ones launched during the observation period alone – but the way these systems leak radiation.

“Some satellites were detected emitting in bands where no signals are supposed to be present at all,” said Dylan Grigg, a Ph.D. candidate at Curtin and lead author of the study, in a press release. “Because they may come from components like onboard electronics and they’re not part of an intentional signal, astronomers can’t easily predict them or filter them out.”

This level of interference will no doubt affect the SKA’s ambitious scientific goals. According to Tingay, the timing of this threat is especially critical. “We’re standing on the edge of a golden era where the SKA will help answer the biggest questions in science: how the first stars formed, what dark matter is, and even test Einstein’s theories,” he stated in a recent press release. Such discoveries depend on near-total radio silence, something astronomers say is now at risk.

A gap in the rules

The issue is not just technical; it’s a regulatory blind spot. According to Tingay, because Starlink is not violating any current rules, it is therefore “doing nothing wrong.” The problem lies with the regulations themselves, which have not kept pace with modern satellite technology.

The International Telecommunication Union (ITU), a United Nations agency, created protected frequency bands decades ago, but its rules focus on intentional transmissions. As Tingay explained, they “do not cover this type of unintended emission.” This regulatory gap means there is currently little policy framework to guide satellite design or hold operators accountable for interference from unintended electromagnetic radiation (UEMR).

The impact of radio interference on astronomy

Unintended emissions are not a new problem, but their scale and intensity are rapidly worsening. As far back as the 1980s, signals from the Soviet-era Global Navigation Satellite System (GLONASS) bled into radio frequencies reserved for science. More recently, in 2006, an Iridium satellite overwhelmed an observation of a faint star even while transmitting in a legally reserved band, and a 2014 study published in IEEE Transactions on Geoscience and Remote Sensing showed how NASA’s soil moisture satellite data has been compromised by ground-based electronics.

But the sheer number of modern satellites has changed the equation. A recent study using the Low-Frequency Array (LOFAR) telescope in the Netherlands found that second-generation Starlink satellites produce UEMR that is 32 times stronger than their predecessors. While SpaceX has made significant progress in mitigating the optical brightness of its satellites to appease astronomers, the radio interference problem appears to be getting worse. 

The source of the leakage remains uncertain. The Curtin paper notes, “Currently, the specific component of the satellite that causes this unintended emission is unknown,” though it is believed to originate from electronics like the propulsion or avionics systems.

Attempts at cooperation 

When it comes to its satellites’ optical brightness, SpaceX has been publicly responsive, developing engineering solutions like sun visors and dielectric mirrors — film that scatters light away from Earth — in response to astronomers’ concerns. 

There have been signs of cooperation on radio interference as well, though the results have been mixed. In 2019, the company signed an agreement with the U.S. National Science Foundation (NSF), which also focused on optical brightness. While that agreement did mention radio interference, the findings from the Curtin and LOFAR studies suggest it has had little impact on the UEMR problem, raising questions about the efficacy of such voluntary measures.

A key example of cooperation on intentional radio signals is the Operational Data Sharing (ODS) system, developed with the National Radio Astronomy Observatory (NRAO).

“This software allows our radio telescopes to share in near-real time where our telescopes are pointed in the sky,” explains NRAO astronomer Christopher De Pree. “This information is utilized by SpaceX to make sure that satellites passing close to our pointing position do not transmit at our observing frequency.” The system is already active at the Very Large Array in New Mexico.

However, systems like ODS are designed to manage intentional transmissions, not the unintended radiation at the heart of the Curtin study. When asked about this distinction, De Pree confirmed that UEMR is a separate and more difficult challenge.

“UEMR is certainly a different issue than signals associated with intended emissions, and also pose greater challenges to radio telescopes that operate more frequently at low frequencies,” he says, noting that these signals can be intermittent and hard to trace. “These types of issues are much easier to identify and solve before the launch of large numbers of satellites.”

Despite the public focus on optical brightness and intended transmissions, the authors of the new study suggest a private dialogue on UEMR is underway. “Discussions we have had with SpaceX on the topic have been constructive,” notes Tingay.

When asked about the specifics of those conversations and how SpaceX is planning to mitigate UEMR, Tingay replies, “It would be best to ask SpaceX representatives.” SpaceX did not respond to requests for comment on its UEMR-mitigation efforts. 

Proposed technical and political solutions

To move forward, astronomers are proposing both technical and political solutions. De Pree’s proposed solution for UEMR focuses on prelaunch prevention through collaboration.

“Careful RFI testing of satellite systems before launch, and the sharing of these tests with radio observatories to give feedback about the detected UEMR, would be a good first step,” he suggests.

This kind of collaborative testing could be facilitated by a facility like the National Radio Dynamic Zone, an idea De Pree proposed in 2023. While not yet a reality, he notes that “there is national funding for experiments to test new technologies and software that can help with this better coordination. However, a dedicated experimental facility is still the long-term goal.”

Tingay says that such a facility would be “a step in the right direction.” He adds his solutions to the mix as well, both political and technical: “Working toward the regulation of UEMR,” “working collaboratively with constellation operators,” “making the public aware,” and “astronomers taking steps to control our own destiny, investing much more in algorithms to remove these signals from our data.”

Filtering unwanted transmissions with algorithms is not a new idea. Development of such systems has been underway for years, progressing from simple digital filters to highly automated software. The latest tools use machine learning to spot and erase unwanted signals.

Ultimately, a lasting solution will require policy change. If the ITU updates its guidelines to cover unintended emissions, future policy could mandate specific shielding requirements, power thresholds for leakage, and standardized testing for all satellites before they leave the ground.

Harmony or interference?

The growing chorus of concern from astronomers highlights a critical need for shared responsibility to ensure that the quest for global connectivity does not silence our view of the cosmos. The conclusion of the Curtin study frames the challenge ahead: “Future mitigation of the UEMR from Starlink satellites will be key for ultra-sensitive experiments with SKA. … SpaceX has made significant progress liaising with astronomers in the optical domain, and we hope to keep this dialogue open in the radio domain.”