🔭 astrophysics

True Masses using RV data with Hipparcos and Gaia Astrometry

This paper presents a model combining Hipparcos and Gaia astrometry with radial velocity data to constrain the orbital inclinations and true masses of long-period companions, revealing that several targets previously classified as brown dwarfs or stars are likely planets.

Original authors: G. Piccinini, A. Petralia, A. Sozzetti, S. Benatti, D. Gandolfi, G. Micela

Published 2026-03-11

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: G. Piccinini, A. Petralia, A. Sozzetti, S. Benatti, D. Gandolfi, G. Micela

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 you are trying to figure out how heavy a mysterious object is, but you can only see it from the side. You can see it wobbling a star back and forth, but you don't know if it's wobbling toward and away from you, or if it's spinning in a circle right in front of your face.

If it's wobbling toward and away, the object is light. If it's spinning in a circle, the object must be incredibly heavy to cause that same amount of wobble. This is the "sin i" problem (where i is the tilt angle) that astronomers have faced for decades: we know the minimum weight of these cosmic objects, but not their true weight.

This paper is like a detective story where the authors finally get a second pair of eyes to solve the mystery.

The Detective Team: Radial Velocity vs. Astrometry

For a long time, astronomers used one tool: Radial Velocity (RV). Think of this like listening to a siren. As a police car (the planet) drives toward you, the pitch goes up; as it drives away, the pitch goes down. By measuring how much the star's "pitch" changes, we know how hard the planet is pulling. But, just like the siren, we don't know the angle of the car. Is it driving straight at you (heavy pull)? Or is it driving past you (light pull)?

The authors added a second tool: Astrometry (specifically data from the Hipparcos and Gaia space missions). If RV is listening, Astrometry is watching. It measures the star's actual position in the sky over time.

The paper combines these two. It's like having a police officer who can both hear the siren and see the car's path on a map. By combining the "sound" (RV) with the "sight" (Astrometry), they can finally calculate the true mass of the object, not just the minimum guess.

The "Smudge" Effect (Proper Motion Anomaly)

The authors use a clever trick called "Proper Motion Anomaly." Imagine you take a photo of a star in 1990 (from the Hipparcos satellite) and another in 2020 (from Gaia).

If the star is alone, it should move in a perfectly straight line across the sky, like a train on a track. But if a heavy planet is tugging on it, the star will wobble off that straight line. The difference between where the star should be and where it actually is is the "anomaly."

The authors built a mathematical model to simulate this wobble. They asked: "If this star has a planet of this mass, at this angle, does the wobble match what we see in the photos?"

The Big Reveal: Planets vs. Brown Dwarfs

For years, many of these long-period companions were classified as Brown Dwarfs (failed stars, too heavy to be planets but too light to be real stars) or even small stars. The authors re-examined nine of these systems and found some surprising plot twists:

The "Heavy" Planets (HD 5388 b & HD 6718 b):
Previous studies thought these were Brown Dwarfs (about 60-80 times the mass of Jupiter). The authors' new model showed they are actually planets, weighing only about 2 to 3 times Jupiter's mass. It turns out they were just tilted in a way that made them look heavier than they were.
The "Light" Brown Dwarf (HD 141937 b):
This one was thought to be a Brown Dwarf. The new analysis suggests it's likely a giant planet, and it might even be tilted so perfectly that it passes directly in front of its star (a transit), which would be a huge discovery for studying its atmosphere.
The "Real" Brown Dwarf (HD 16760 b):
This one really is a Brown Dwarf (about 20 times Jupiter's mass). But the authors found a bonus: there might be a second, smaller planet hiding in the same system, like a dog chasing a cat in the same yard.
The "Complicated" Family (30 Ari B b):
This system is messy. It has a massive companion (a small star), but the authors found that the "wobble" they were measuring was actually caused by a third star far away in the system, not the planet they were studying. It's like trying to weigh a baby while someone else is shaking the crib.

Why Does This Matter?

Think of the "Brown Dwarf Desert." Astronomers used to think there were very few objects in the "Goldilocks zone" between planets and stars. But if we've been misidentifying planets as Brown Dwarfs because we didn't know their tilt, our map of the universe is wrong.

By getting the true mass, the authors are:

Correcting the census: Realizing there are more giant planets and fewer "failed stars" than we thought.
Understanding formation: Knowing the true mass helps us understand how these solar systems were built. Did the giant planet form far out and move in? Or did it form right there?
Finding the hidden: They are proving that by combining old data (Hipparcos) with new data (Gaia), we can see things we missed before.

The Bottom Line

This paper is a masterclass in "connecting the dots." By taking 30 years of star-watching data and combining it with modern super-precise measurements, the authors have turned a bunch of "maybe-massive" objects into a clearer picture of our cosmic neighborhood. They showed us that sometimes, the heavy things aren't as heavy as they look, and the light things are exactly what they seem to be.

1. Problem Statement

The fundamental challenge in exoplanet characterization is determining the true mass ( $M_c$ ) of a companion. Radial Velocity (RV) surveys, while effective at detecting long-period companions, only provide a minimum mass ( $M_c \sin i$ ) because the orbital inclination ( $i$ ) relative to the observer is unknown. This ambiguity leads to significant classification errors:

Objects with $M_c \sin i < 13 M_{Jup}$ (the deuterium-burning limit) could be massive planets, brown dwarfs (BDs), or low-mass stars if the inclination is low (face-on).
Previous studies relying solely on RV data often overestimate the occurrence rate of giant planets because they cannot distinguish between planetary-mass objects and stellar/BD contaminants.
Previous attempts to resolve this using Gaia Data Release 1 (DR1) astrometric excess noise (e.g., Kiefer et al. 2021) suffered from limitations, including short time baselines (418 days) and the inability to model multiple companions or complex orbital dynamics accurately.

2. Methodology

The authors developed a joint modeling framework that simultaneously fits RV data and astrometric Proper Motion Anomalies (PMA) to break the $\sin i$ degeneracy.

Data Sources:
- RV Data: Long-baseline spectroscopic data from various instruments (HARPS, HIRES, SOPHIE, etc.), covering decades of observation.
- Astrometry: Combined data from Hipparcos (epoch ~1991) and Gaia (DR2/DR3, epochs ~2015–2016). The ~25-year baseline allows for the detection of long-term acceleration.
The PMA Technique:
- The model utilizes the difference between the long-term proper motion (Hipparcos-Gaia baseline) and the short-term proper motion (Gaia or Hipparcos alone).
- This difference ( $\Delta \mu$ ) represents the acceleration of the photocenter due to an orbiting companion.
- The model accounts for the "smearing effect" (time-averaging of proper motion over the mission duration) which reduces sensitivity for periods comparable to the mission baseline.
Statistical Approach:
- Simultaneous Fitting: The authors use a Differential Evolution Markov Chain Monte Carlo (DE-MCMC) algorithm to jointly fit the Keplerian RV parameters ( $P, T_p, e, \omega, K$ ) and astrometric parameters ( $i, \Omega$ ).
- Likelihood Function: Combines the $\chi^2$ of the PMA residuals and the RV residuals (including instrumental offsets and jitter terms).
- Priors: Informative uniform priors were used to avoid bias from previous literature, exploring both prograde ( $i < 90^\circ$ ) and retrograde ( $i > 90^\circ$ ) solutions.
Validation Tools:
- Sensitivity Curve: A theoretical tool (based on Kervella et al. 2019) used to verify if the derived mass and semi-major axis are consistent with the observed astrometric signal, accounting for orbital period and observational windows.
- RUWE (Renormalized Unit Weight Error): Used as a quality check for the Gaia single-star solution; high RUWE values suggest unresolved companions.

3. Key Contributions

Methodological Validation: The code was validated against three benchmark systems (GJ 463 b, $\pi$ Men b, HD 222237 b) with known literature results, demonstrating consistency in derived masses and inclinations.
Re-analysis of Kiefer et al. (2021): The authors re-examined 8 systems previously analyzed using Gaia DR1 and the GASTON tool, applying the superior joint RV+PMA modeling with Gaia DR3 data.
Resolution of Mass Degeneracy: Successfully constrained the true masses of several long-period companions, reclassifying objects previously thought to be brown dwarfs or stars as planets, and vice versa.
Detection of Multi-Companion Systems: Identified residual signals in RV data suggesting the presence of additional companions in systems like HD 5388 and HD 16760.

4. Key Results

The study analyzed 8 targets, yielding the following significant findings:

Reclassification from BD/Star to Planet:
- HD 5388 b: Previously classified as a brown dwarf ( $\sim 87 M_{Jup}$ ) or M-dwarf. The new analysis yields a true mass of $3.2^{+0.3}_{-0.2} M_{Jup}$ , confirming it as a planet. A residual signal suggests a possible second sub-Jovian companion.
- HD 6718 b: Previously estimated at $\sim 63 M_{Jup}$ (BD regime). The new mass is $2.4^{+0.3}_{-0.2} M_{Jup}$ , classifying it as a planet.
- HD 141937 b: Previously estimated at $\sim 27 M_{Jup}$ (BD regime). The new analysis finds an edge-on orbit ( $i \approx 90^\circ$ ) and a mass of $11.3 \pm 0.5 M_{Jup}$ , firmly placing it in the giant planet regime. The authors note it could be a transiting planet, though TESS did not detect a transit due to the short observation window and narrow geometric probability.
Brown Dwarf Confirmation:
- HD 16760 b: Confirmed as a brown dwarf with a mass of $\sim 21.5 M_{Jup}$ . The system shows signs of a second companion (likely a planet) and is part of a wider visual binary system, making it a complex target for formation models.
Stellar Companions & Complex Systems:
- 30 Ari B b: Confirmed as a low-mass star ( $\sim 188 M_{Jup}$ ). The study recovered the signal of a tertiary stellar companion (30 Ari C), which significantly impacts the PMA signal and complicates the characterization of the inner companion.
- HD 148427 b, HD 96127 b, HIP 65891 b: For these three systems, the orbital inclination could not be tightly constrained (ranges of $50^\circ$ – $130^\circ$ ). However, the derived masses ( $\sim 1.2$ , $3.8$, and $6.1 M_{Jup}$ respectively) are lower than previous stellar classifications and consistent with substellar objects, supported by low RUWE values.
Discrepancy with GASTON/DR1:
- The masses derived in this work are consistently lower than those from Kiefer et al. (2021). The authors attribute this to the limitations of the GASTON tool (simulating noise rather than modeling dynamics) and the short Gaia DR1 baseline, which failed to capture full orbital phases for long-period targets.

5. Significance

Correcting Occurrence Rates: By reclassifying several "brown dwarfs" as planets, this study suggests that the true occurrence rate of long-period giant planets may be higher than previously estimated when using minimum masses alone.
Methodological Advancement: The joint modeling of RV and PMA, validated against the sensitivity curve, provides a robust, independent method for determining true masses without relying on simulated excess noise.
System Complexity: The detection of additional companions in systems like HD 5388 and HD 16760 highlights the prevalence of multi-body systems, which are crucial for understanding planet formation and migration (e.g., the role of giant planets in inhibiting or promoting inner planet formation).
Future Outlook: The authors emphasize the need for Gaia DR4 data to further refine these constraints, particularly for systems where the current inclination ranges are still broad. The integration of Gaussian Processes for stellar activity is noted as a future step to further refine these models.