Space & astronomyPreprintObservation4 min read

A FAMOUS BLACK HOLE GOES QUIET

Most black holes in binary systems spend their lives asleep. A stellar-mass black hole pulls gas from a companion star into a disc, and the disc stores it. Every so often, the stored gas pours in: the system erupts for a few months, shining in X-rays and firing jets, then falls quiet again.

GRS 1915+105 never followed that script. Discovered at the start of an eruption in 1992, it stayed extremely bright for about 25 years and active for decades. It became the archetypal microquasar of our Galaxy, known for its recurring jets and repeated ejections of matter. Its black hole weighs about 12.4 times the Sun, it lies 9.4 kiloparsecs away (the kiloparsec is the unit astronomers use for distances across the Galaxy), and its companion takes 33.85 days to orbit: the longest orbit of any known X-ray binary, and so the largest disc.

Hidden, then fading

In mid-2019, its X-rays dropped sharply. The source had become wrapped in a thick screen of gas, so thick that X-rays could hardly escape. Its radio emission, which passes through that screen, stayed bright, at times flaring above one jansky (the unit radio astronomers use to measure how bright a source appears). So the engine was still running, just out of sight.

Sara Motta, of the Brera Astronomical Observatory (INAF) in Italy, R. P. Fender, of the University of Oxford, and colleagues across Europe, China, Chile, Canada, the United States and South Africa have now followed what happened next. They used 155 observations from the South African radio telescope MeerKAT since December 2018, the AMI-LA radio telescope, and several X-ray satellites.

The X-ray monitors lost the source for good in 2025: the last detection in lower-energy (“soft”) X-rays came in April, the last in higher-energy (“hard”) X-rays in September. But X-rays alone cannot tell a dimmer engine from a thicker screen. The radio can.

After a last major flare of more than one jansky in early 2025, the radio signal sank steadily: about 1.1 millijansky in October 2025, 0.55 in January 2026, then just a few tenths. After 9 May 2026, the short routine observations no longer detected it.

The faintest glow on record

In July 2026, the team pointed MeerKAT at the source for about eight hours in each of two frequency bands. It was still there: 150 microjanskys at 1.28 gigahertz and 93 at 2.6 gigahertz — the lowest radio levels ever recorded from GRS 1915+105, about ten times below the level of late 2025. The signal did not vary over six days.

For the first time, the source is faint in every band in which it has been observed. Earlier quiet spells had always been partial: faint in radio while bright in X-rays, or the reverse.

The new radio luminosity is exactly what a quiet black-hole binary with such a long orbit should emit: the relation traced by 17 other quiet systems, most of them with upper limits only, predicts a value only 4% below the measurement. It also makes GRS 1915+105 the most radio-luminous quiet black-hole binary known, about ten times brighter than the previous record holder, V404 Cygni. The authors attribute this to its giant companion, which sheds gas about a hundred to a thousand times faster than the companions in tighter systems, and to its huge disc.

Not just a thicker screen

Could the engine still be roaring behind an even denser screen? The authors consider it unlikely. To hide an undimmed engine from a deep observation in late 2025, the amount of gas along our line of sight would have had to grow more than tenfold. Such a screen would also re-emit the absorbed energy as infrared light, which the SPHEREx infrared survey did not detect. The radio decline, they write, is the first direct evidence that the central engine itself is fading.

One detail does not fit the textbook. The radio spectrum falls steeply with frequency, while the jets of quiet black holes usually show a flat spectrum. The most economical reading, according to the authors, is that leftover matter ejected during the 2025–2026 activity still glows faintly. If so, the jet at the heart of the system is fainter still — which only strengthens the case for quiet.

Ten thousand years of calm

Models of the disc had predicted that this eruption would last about 20 to 100 years before the gas reservoir ran low: 34 years have passed. They also predict a quiet phase on the order of 10,000 years, so the transition under way will not repeat on any timescale astronomers can watch.

The silence opens two doors. The giant companion star can be studied without the glare of the disc. And the very-high-energy gamma rays detected from this system become a test: if they keep shining at the same level, they must come from where the jets strike the surrounding gas, tens of parsecs from the black hole; if they fade within a few years, part of them comes from close to it.