Starlink signal leakage up to 10,000 times stronger than SKA-Low's cosmic targets

A Curtin University study has catalogued more than 112,000 unintended radio emissions from Starlink satellites sitting squarely in the frequency range the Square Kilometre Array Low (SKA-Low) observatory depends on, with the leakage reported to be up to 10,000 times stronger than the cosmic signals the telescope exists to detect. According to Gadget Review, the work was carried out using a prototype SKA-Low station in Australia and published in the journal Astronomy & Astrophysics.

What the study measured

The Curtin team used the Engineering Development Array 2, a prototype station for SKA-Low, to analyse roughly 76 million radio images gathered over 29 days. Within that data they identified 112,534 individual emissions coming from 1,806 distinct Starlink satellites across 73–235 MHz — the exact window in which SKA-Low operates.

Some of the emissions follow a pattern: periodic tones spaced 13 kHz apart around 137 MHz, repeating approximately every 100 seconds. Crucially, these are not scheduled downlinks that observatories can plan around. They are accidental radiation from onboard electronics, and because the emissions are unpredictable, astronomers cannot model and subtract them from data the way they handle known, repeating signals.

Two figures frame the severity. The measured leakage reaches up to 10⁶ Jy per beam, while the neutral hydrogen signals SKA-Low is hunting demand sensitivity on the order of 10⁻⁵ Jy. Steven Tingay of Curtin University described the leaked emissions as "comparable to the brightest natural radio sources in the sky." Gadget Review reports the Starlink signals run roughly 10,000 times stronger than what SKA-Low is built to detect, and at some frequencies up to 30% of the collected images contained Starlink interference.

The emissions also appeared inside two bands the International Telecommunication Union (ITU) reserves for radio astronomy — 73–74.6 MHz and 150.05–153 MHz — spectrum where such signals are not supposed to exist at all.

A regulatory blind spot

Structurally, nobody is breaking the rules. ITU protections cover intentional transmissions into protected astronomy bands. Unintended electromagnetic radiation — hardware leakage — sits largely outside that framework, meaning it remains effectively unregulated even when it bleeds into reserved spectrum. The problem compounds with scale, since the Starlink constellation already exceeds 6,000 satellites and continues to grow.

What could fix it

SpaceX has cooperated with astronomers before. Sunshades reduced the optical brightness of its satellites, and a coordination agreement with the US National Science Foundation addressed interference at higher radio frequencies. Neither measure touches low-frequency hardware leakage, which is baked into the electronics and structure of the satellites themselves.

According to Gadget Review, the researchers shared their findings with SpaceX, which is reportedly open to discussing hardware changes in future satellite designs. Algorithmic mitigation — filtering the interference out in software — is being explored, but scientists describe that work as "embryonic," and it could demand computing resources comparable to the scientific processing itself. The researchers argue that engineering changes to the satellites are the real solution, just as design changes resolved the optical brightness problem.

Why it matters

SKA-Low was designed to detect neutral hydrogen from the cosmic dawn roughly 13 billion years ago, when the first stars ignited — among the faintest observations ever attempted. Interference at this scale does not merely degrade data quality; it threatens to make entire frequency bands scientifically unusable. Because the leakage falls into a regulatory gap, any resolution depends on voluntary cooperation rather than enforceable rules, and each additional satellite launched widens the contamination. The study offers an early, concrete measurement of a widening trade-off between satellite mega-constellations and ground-based science: the usable radio sky is a finite resource, and the decision about how to protect it may need to be made before the window for these observations closes.