Revision #1

A Stellar Wind Reveals the Stopwatch on a Black Hole's Jet

For half a century, astronomers have known that some black holes launch twin beams of plasma — relativistic jets — that travel across light-years and sculpt the galaxies around them. What they have not been able to do is put a stopwatch on one. Every previous estimate of jet power averaged energy deposited into surrounding gas over tens of thousands of years, more archaeology than measurement [1][2].

That changed on April 16, 2026, when a team led by Dr. Steve Prabu, now at the University of Oxford and previously at Curtin University's node of the International Centre for Radio Astronomy Research, published in Nature Astronomy the first direct, near-instantaneous measurement of a black hole jet's kinetic power and bulk speed [3][4]. The target was Cygnus X-1 — the first object ever identified as a black hole, spotted in 1964 — and the trick was using the howl of its companion star's wind as a natural anemometer.

What was measured

The team clocked the jet from Cygnus X-1 at roughly half the speed of light, or about 150,000 kilometers per second (540 million kilometers per hour) [1][2]. In the paper's own units, the logarithm of the jet's kinetic power was 37.3 (in erg s⁻¹), which works out to roughly 2 × 10³⁰ watts — equivalent to the combined luminous output of about 10,000 Suns [3][5]. The reported uncertainty band is tight: +0.1 dex upward, -0.2 dex downward [3].

Source: Compiled from published measurements

Data as of Apr 16, 2026CSV

Cygnus X-1 itself is a binary system about 7,200 light-years away in the constellation Cygnus, consisting of a black hole estimated at roughly 21 solar masses orbiting the blue supergiant HD 226868 every 5.6 days at a separation of about 0.2 astronomical units [6][7]. The star's fierce radiation-driven wind pours past the black hole, and some of that wind is funneled onto the hole to power the X-ray emission that first gave Cygnus X-1 its name [6].

Critically, the new jet-speed figure — around 0.5c, implying a bulk Lorentz factor of roughly 1.15 — sits at the low end of what theorists have long guessed for stellar-mass black hole jets and well below the Lorentz factors of 10 to 30 inferred for jets from supermassive black holes in blazars [8]. Recent statistical work published in Monthly Notices of the Royal Astronomical Society argues that the underlying Lorentz-factor distributions for stellar-mass and supermassive black hole jets are actually consistent once selection effects are accounted for — a point the Cygnus X-1 measurement helps anchor with a real number rather than an inferred one [8].

Source: Compiled from VLBI observations

Data as of Apr 16, 2026CSV

How 18 years of radio images became a speedometer

The technical innovation is elegant. Between 2002 and 2020, Prabu's team and collaborators accumulated high-resolution radio images of Cygnus X-1 using Very Long Baseline Interferometry (VLBI) — a technique that links radio telescopes separated by thousands of kilometers to synthesize an effective aperture the size of a continent, with angular resolution finer than a ten-thousandth of an arcsecond [9].

In those images, the jet does not stand still. As the black hole swings around its 5.6-day orbit, the blue supergiant's wind buffets the outflow from different angles, bending the jet visibly like a hose sprayed by a crosswind [5]. Because the stellar wind's mass-loss rate and velocity are independently measurable — HD 226868 is a well-studied O-type supergiant with a surface temperature near 31,000 K — the researchers could treat the jet's deflection as a balance of momentum flux between jet and wind [6][5]. Knowing the push, they could back out the jet's carried momentum, and therefore its power and speed [1][5].

Three conditions had to line up before this measurement was possible. First, accumulated baselines had to exceed a decade so the jet's repeated bending pattern could be pinned down. Second, independent modeling of the stellar wind from the O-type companion had to mature to the point where its thrust could be trusted as a reference. Third, the computer simulations of jet-wind hydrodynamics had to be good enough to convert observed deflections into physical parameters. The team credited the 18-year baseline and dedicated hydrodynamic modeling as the elements that made the calculation possible now rather than a decade ago [1][3][5].

The energy budget: 10% goes into the jet

Perhaps the most consequential number in the paper is not the speed but the ratio. The measured jet kinetic power is "comparable to the accretion energy determined from its bolometric X-ray luminosity" — about 10% of the energy released by matter falling onto the black hole is carried out of the system in the jet, with the remainder radiated as X-rays and other electromagnetic emission [3][1][5].

Source: Prabu et al., Nature Astronomy 2026

Data as of Apr 16, 2026CSV

"This is what scientists usually assume in large-scale simulated models of the Universe, but it has been hard to confirm by observation until now," Prabu said in a statement accompanying the paper [5]. That assumption shows up in cosmological simulations including IllustrisTNG, XFABLE, and the COLIBRE project — all of which insert an efficiency factor for how much of an accreting black hole's energy gets converted into mechanical "kinetic-mode" feedback that heats surrounding gas and regulates star formation in massive galaxies [10][11].

The relevance to jet-launching theory is direct. The Blandford–Znajek process, proposed in 1977, posits that jets tap the rotational energy of a spinning Kerr black hole through magnetic field lines anchored in the surrounding accretion disk [12]. In that picture, the jet power scales with the square of the black hole's spin parameter and with the magnetic flux threading the event horizon [12]. A measured efficiency of roughly 10% for Cygnus X-1 — a system known to host a rapidly spinning black hole — is in the range Blandford–Znajek simulations predict, but it does not by itself prove the mechanism. General-relativistic magnetohydrodynamic (GRMHD) simulations have reproduced similar efficiencies in several competing scenarios [12].

How much is 10,000 Suns, really?

In absolute terms, the Cygnus X-1 jet is a modest engine on the cosmic scale. The Milky Way's total stellar luminosity is around 1.3 × 10³⁷ watts — roughly seven orders of magnitude above the jet's output [13]. A single luminous quasar like 3C 273 hosts jets estimated near 10⁴⁰ watts, ten million times more powerful still [10]. The brightest gamma-ray burst ever recorded, GRB 221009A, hit a peak isotropic-equivalent luminosity of about 2.1 × 10⁴⁷ watts — seventeen orders of magnitude above Cygnus X-1's jet [14].

Why does a 10,000-Sun jet matter, then? Because Cygnus X-1 is the nearest stellar-mass analog of the feedback engines that dominate galaxy evolution, and because the same physics — 10% of accretion energy shunted into kinetic power rather than light — is what simulations assume operates in the supermassive black holes at the centers of galaxies when they cycle through their "radio-mode" feedback state [10][11]. Observing the mechanism in miniature, with a resolvable system only 7,200 light-years away, gives theorists a benchmark they did not previously have.

How generalizable is a single measurement?

Cygnus X-1 is one system. The BlackCAT catalog lists 59 confirmed or candidate stellar-mass black holes in Galactic X-ray binaries, and the broader population of jetted active galactic nuclei numbers in the thousands [15]. The study's framing as the "first-ever" jet power and speed measurement is carefully narrow: it refers to a direct, near-instantaneous kinematic measurement, not to prior indirect or time-averaged estimates that already existed for many sources [1][2].

Whether the result generalizes is an open question. Cygnus X-1 is a high-mass X-ray binary in a "hard" spectral state where its jet is persistent and compact — a regime distinct from the transient "ballistic" jets launched during outbursts of low-mass X-ray binaries like GRS 1915+105, whose ejecta have been clocked at apparent superluminal speeds corresponding to true velocities above 0.9c [8]. Whether the 10% efficiency and 0.5c speed scale up to supermassive black holes, whose jets routinely show Lorentz factors above 10, remains a theoretical extrapolation. A March 2026 paper in The Astrophysical Journal on the M87 jet reported apparent transverse wave propagation at 2.7–2.9 times the speed of light — a clear superluminal effect due to geometry rather than a genuine speed, but a reminder that supermassive-black-hole jets occupy a very different dynamical regime [16].

Where the skeptics press

The methodology's strongest pressure points are the assumptions it inherits from its ancillary inputs. Three stand out.

First, the geometry of Cygnus X-1 is not nailed down. Published estimates of the system's orbital inclination range from 27° to 65°, with the jet axis itself estimated at around 30° to the line of sight but likely precessing [17][6]. A 2022 Science study using X-ray polarimetry concluded that the accretion disk is viewed closer to edge-on than the binary orbit — a misalignment that, if severe, alters the projection corrections used to convert observed jet bends into intrinsic velocities [17]. Because the derived jet speed depends on the Doppler factor and thus the inclination, a shift in viewing angle can in principle move the inferred power by a substantial factor.

Second, the stellar wind from HD 226868 is a modeled quantity rather than a direct observable. Published mass-loss rates for O-type supergiants carry a factor-of-two-or-more uncertainty, and the wind's structure near the black hole is distorted by the compact object's gravity and photoionization. If the effective ram pressure deflecting the jet is off by a factor of two, so is the inferred jet momentum.

Third, the conversion from momentum flux to kinetic power assumes the jet's composition — the ratio of protons to electron-positron pairs and the partitioning between bulk motion and internal energy — is known. The paper's quoted uncertainty of ±0.1 to 0.2 dex on jet luminosity captures statistical errors and some modeling variance, but composition priors propagate into the final number in ways that are partly philosophical [3].

None of these concerns invalidates the result; they circumscribe its precision. A speed of "roughly half the speed of light" is robust to the order-of-magnitude level. Whether the true figure is 0.4c or 0.6c — and whether the efficiency is 5% or 20% — will depend on follow-up work with denser sampling and better-constrained system parameters.

Why cosmological simulators are paying attention

Black hole feedback is the load-bearing assumption in modern galaxy formation theory. Without it, cosmological simulations overproduce massive galaxies by factors of several, fail to quench star formation in massive halos, and miss the observed galaxy luminosity function at the bright end [10][11]. The energy injected by supermassive black holes into their surroundings has been estimated to exceed the energy budget of supernovae by a factor of 20 to 50 in those galaxies [10].

Simulations fine-tune a parameter — the efficiency with which accreted rest-mass energy converts into mechanical outflow — and calibrate it to match galaxy statistics. That parameter has been essentially unconstrained by direct observation at the level of individual systems. An empirical 10% number from Cygnus X-1 does not end the debate, but it provides a data point where previously there were only theoretical priors [3][5][10].

Source: OpenAlex

Data as of Jan 1, 2026CSV

Upcoming facilities should multiply such measurements. The Square Kilometre Array, whose "Array Assembly 2" milestone is slated for 2026/2027 and whose full science operations are expected by 2028, will have the sensitivity to track jet structure in dozens of Galactic X-ray binaries and in nearby active galactic nuclei at cadences VLBI arrays cannot currently match [18]. Research on black hole jets has been a publication focus since the 2010s — OpenAlex records more than 51,000 papers on the topic to date, with a peak of 6,210 published in 2023 — and each new direct measurement narrows the gap between simulation assumptions and observational ground truth [19].

The verdict astronomers are converging on

The Prabu measurement does three things at once. It confirms, to roughly a factor of two, the efficiency assumption that cosmological simulators have baked into their AGN-feedback prescriptions for more than a decade. It demonstrates that a jet from a relatively ordinary microquasar carries kinetic power on the order of 10,000 Suns, a figure that until now was bracketed only by inference. And it opens a methodological door: wind-deflected jets give astronomers a natural probe wherever a black hole orbits a luminous companion — a configuration that applies to dozens of known X-ray binaries in the Milky Way alone [5][15].

Prior indirect estimates relied on measuring the bubble of heated gas carved out of the surroundings by a jet over its lifetime — an approach that works for the largest radio lobes in galaxy clusters but smears over orbital variability, transient outbursts, and changes in jet composition [1][2]. Watching a jet bend in real time, and knowing what is doing the bending, is a substantively different kind of observation.

It does not settle whether Blandford–Znajek is the dominant launching mechanism, nor whether Cygnus X-1's efficiency maps directly onto a quasar ten light-years across. What it does provide is a single, hard number in a field that has run on soft numbers for decades.