A Dance of Dust and Light Around a Young Star express
460 light-years away, around the star HD 135344B, a dense disk of gas and dust is swirling. It looks like a giant vinyl record that someone has cracked: a gap yawns in the disk, and across it, like a slender bridge, stretches a dust filament. Next to this filament, in scattered light, a twist is visible — a bend in the spiral arm, as if it had been briefly held back and then released.
Astronomers gathered images from four instruments (VLT/NACO, VLT/SPHERE, VLT/ERIS, and JWST/NIRCam) over ten years. They applied algorithms that remove the glare from the bright star without distorting extended structures. It turned out that the twist moves together with the spiral at a rate of 0.81±0.05 degrees per year. This nearly matches earlier measurements of 0.85±0.05 degrees per year.
Such motion is naturally explained if a giant planet lurks in the gap. Its orbital speed at a distance of 69±4 astronomical units (twice as far as Neptune is from the Sun) matches the observed shift of the twist. A recent image that seemed to show the planet itself turned out to be just an artifact — a ghost born from the algorithm. But the twist, the spiral, the gap, and the dust filament all fit into a single pattern drawn by an invisible hand.
Original abstract
as written by the authors arXiv:2606.10624
The HD 135344B (SAO 206462) disk exhibits strong signposts of planet formation. ALMA images in the sub-mm revealed a gap-crossing dust filament whose position coincides with a twist detected in the scattered-light spiral structure. Analysis of the spirals in polarized light also hints at a spiral-driving protoplanet in the sub-mm gap. We aim to study the spirals dynamics, as well as the twist, over a 10-year baseline, in different bands. We also seek to assess the authenticity of a recently claimed candidate protoplanet. We use high-fidelity post-processing algorithms such as IPCA to minimize the biases induced by ADI on extended sources and analyze archival VLT/NACO, VLT/SPHERE, VLT/ERIS and JWST/NIRCam datasets to obtain the spiral traces and measure their orbital motion in multiple scattered light bands. We measure an average spiral orbital motion of 0.81$$\pm$$0.05 deg/yr, in agreement with the literature value of about 0.85$$\pm$$0.05 deg/yr at all wavelengths. With simple modeling of the twist morphology, we confirm that it is co-moving with the spiral in which it is embedded. While the position angle of the twist coincides with the dust filament, it is located at a smaller angular separation from the star, which we attribute to the fact that the spiral trace moves away from the central star with increasing wavelength. We find that a recently claimed protoplanet candidate can be explained as a post-processing artifact. Our confirmation that the motion of the scattered light twist is consistent with the orbital velocity of a planet at 69$$\pm$$4 au over a 10-year baseline suggests that the spirals, the gap, the dust filament, and the twist, could indeed be attributed to the same hypothetical protoplanet embedded within the spiral. A perplexing trend for a wavelength-dependence of the angular distance of the spiral traces to the central star remains to be explained.
One astronomical unit is the distance from Earth to the Sun. The invisible planet in the disk of HD 135344B lies 69 times farther from its star than Earth is from the Sun, meaning it would be almost at the edge of our solar system if placed here.
In Arthur C. Clarke's novel "Rendezvous with Rama," humans explore a mysterious object that hides its secrets behind a wall. So here: the planet remains invisible, but its presence is betrayed by a gravitational signature in dust and light.