🔭 astrophysics

HD 38973 b -- a cold Saturn orbiting a Sun-like star

Using precision radial-velocity measurements combined with astrometric constraints from the Hipparcos–Gaia Catalog of Accelerations, researchers detected a cold Saturn-mass companion orbiting the Sun-like star HD 38973 with a period of approximately 2733 days, demonstrating how non-detections in astrometry can provide robust upper limits on true companion masses.

Original authors: Adriana Errico, Robert A. Wittenmyer, Jonathan Horner, Brad Carter, Alexander Wallace

Published 2026-05-21

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Adriana Errico, Robert A. Wittenmyer, Jonathan Horner, Brad Carter, Alexander Wallace

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine the night sky as a giant, quiet dance floor. For decades, astronomers have been trying to spot the dancers (planets) by watching how the music (stars) sways. Usually, we can only see the dance from the side, which makes it hard to tell how heavy the dancers are or exactly how they are moving.

This paper is about a new discovery on that dance floor: a planet named HD 38973 b orbiting a star that looks very much like our own Sun. Here is the story of how they found it and what they learned, explained simply.

The "Wobble" in the Star

The astronomers used a technique called Radial Velocity. Think of the star as a heavy person on a swing, and the planet as a small child pushing them. Even though the child is small, their pushes make the heavy person sway back and forth.

By measuring the star's "wobble" with incredibly precise instruments (like a super-accurate ruler for speed), the team saw a pattern. The star was swaying back and forth in a rhythm that took about 7.5 years to complete one full cycle. This long, slow rhythm told them there was a planet out there, but because they were only looking from the side, they couldn't be sure exactly how big the planet was. They knew it was at least as heavy as Saturn, but it could have been much heavier if the orbit was tilted just right.

The "Ghost" Signal

To get a better look, the team tried to use Astrometry. If you imagine the star is a lighthouse, astrometry tries to see if the lighthouse beam is wobbling in the sky from our perspective.

However, for this specific planet, the wobble was too tiny to see directly. It was like trying to see a single grain of sand move on a beach from a satellite. The instruments (Hipparcos and Gaia) didn't see a clear wobble.

Here is the clever part: Even though they didn't see the wobble, the fact that they didn't see a big wobble told them something important.

If the planet were a giant, heavy beast (like a brown dwarf or a small star), it would have caused a huge, obvious wobble that the instruments would have seen.
Since the instruments saw nothing, the planet cannot be a giant. It must be relatively small.

It's like walking into a room and not hearing a drum. You can't tell exactly how loud the drum is, but you know for sure it isn't a massive bass drum shaking the walls.

Putting the Clues Together

The team combined the two pieces of evidence:

The Side View (Radial Velocity): "The star is wobbling with a rhythm that suggests a Saturn-sized object, but it could be heavier if the orbit is tilted."
The Top View (Astrometry): "We didn't see a big wobble, so the object can't be too heavy."

By mixing these clues, they solved the mystery. They calculated that the planet is almost certainly a "Cold Saturn."

Cold: It orbits far away from its star (about 4 times the distance from Earth to the Sun), so it's chilly.
Saturn-sized: It has a mass of about 0.24 times the mass of Jupiter (which is roughly the mass of Saturn).

Why This Matters

This paper highlights a new way of doing detective work in space. Usually, if you can't see a signal (like the wobble), you might give up. But this study shows that not seeing something can be just as powerful as seeing it. It helps rule out the "scary" possibilities (like a hidden star) and confirms that we are looking at a genuine, friendly planet.

In short: The astronomers found a Saturn-like planet orbiting a Sun-like star. They couldn't see the planet directly, and the star's wobble was too faint to measure perfectly, but by proving the wobble wasn't big enough to be a giant monster, they confirmed it's a planet we can be excited about.

Technical Summary: HD 38973 b - a cold Saturn orbiting a Sun-like star

Problem Statement
The detection and characterization of long-period exoplanets, specifically "Jupiter analogues" or systems resembling our own Solar System, remain challenging. Radial velocity (RV) surveys can detect massive planets on long-period orbits, but this technique suffers from a fundamental mass-inclination degeneracy, yielding only a minimum mass ( $M_p \sin i$ ) rather than the true mass. Furthermore, the astrometric signal induced by such planets is often below the detection threshold of current space-based observatories like Gaia and Hipparcos, leading to non-detections that are traditionally viewed as uninformative. This study addresses the problem of constraining the true mass and orbital inclination of a long-period companion when the astrometric signal is not directly detected.

Methodology
The authors present a multi-faceted approach combining precision radial velocity measurements with astrometric constraints from the Hipparcos–Gaia Catalog of Accelerations (HGCA).

Radial Velocity Analysis: The study utilizes 48 epochs of RV data from the Anglo-Australian Telescope (AAT/UCLES) spanning 1998–2015 and 45 epochs from the HARPS spectrograph spanning 2003–2012. The data were analyzed using a Generalised Lomb-Scargle (GLS) periodogram, which revealed a highly significant signal ( $FAP \approx 6 \times 10^{-8}$ ) near 3000 days. Orbital parameters were modeled using a single-planet Keplerian fit via the RadVel package, employing Markov Chain Monte Carlo (MCMC) sampling. Stellar activity indicators (FWHM, BIS, $\log R'_{HK}$ , and Contrast) were analyzed to rule out magnetic cycles as the source of the signal.
Astrometric Constraints: The authors calculated the expected astrometric reflex motion based on the RV-derived parameters, finding it to be $\sim0.027$ mas, which is below the single-epoch precision of Hipparcos and Gaia. Consequently, they focused on the proper-motion anomaly (the difference between short-term Gaia proper motion and the long-term Hipparcos–Gaia baseline). The observed anomaly for HD 38973 was found to be consistent with zero.
Joint Inference Strategy: To resolve the mass-inclination degeneracy, the authors employed a joint inference strategy. They generated a distribution of true companion masses by marginalizing over the unknown orbital inclination (assuming an isotropic distribution) using the RV posterior samples. They then applied the HGCA likelihood—derived from the non-detection of a proper-motion anomaly—to weight these samples. This process effectively downweighted high-mass, low-inclination solutions that would have produced a detectable astrometric signal inconsistent with the HGCA data.

Key Results

Detection: A coherent Keplerian signal was confirmed with a period of $2733^{+210}_{-190}$ days and a moderate eccentricity of $0.23 \pm 0.15$ .
Minimum Mass: The RV analysis alone yields a minimum mass of $M_p \sin i = 0.22 \pm 0.03 M_{Jup}$ .
True Mass and Inclination: By combining the RV posterior with the HGCA likelihood, the study derives a robust true mass of $0.240^{+0.102}_{-0.040} M_{Jup}$ . The analysis places a 95% confidence lower limit on $\sin i$ of 0.5, constraining the orbital inclination to between $27^\circ$ and $153^\circ$ and ruling out near-face-on configurations.
Mass Limits: The joint analysis establishes a 95% upper limit on the companion mass of $0.484 M_{Jup}$ , significantly tighter than the RV-only limit of $0.662 M_{Jup}$ . This confirms the companion is a planet and rules out stellar or brown dwarf companions.
Stellar Activity: No significant correlations were found between the RV signal and stellar activity indicators, supporting the planetary nature of the signal.

Significance and Claims
The paper claims that the primary significance of this work lies in demonstrating the "diagnostic power of astrometric non-detections." The authors argue that even when an astrometric signal is too weak to be directly detected, the absence of such a signal provides informative upper limits on companion mass. By combining precision RVs with HGCA likelihoods, it is possible to rule out high-mass solutions at low inclinations and derive robust constraints on the true mass of long-period companions.

The study identifies HD 38973 b as a likely "cold Saturn" orbiting a Sun-like star at a semi-major axis of $\sim3.91$ AU. Additionally, the authors note that constraining the inclination slightly increases the calculated transit probability from $\sim0.15\%$ (assuming isotropic inclinations) to $0.20\%$ , illustrating how combined RV and astrometric analyses can help identify systems with a realistic potential for future transit observations. The work suggests that as Gaia astrometry improves, this methodology will become increasingly valuable for characterizing long-period planets.

The "Wobble" in the Star

The "Ghost" Signal

Putting the Clues Together

Why This Matters

More like this