Predicting Gaia astrometry's ability to constrain the populations of circumbinary planets

Imagine the universe as a giant, crowded dance floor. For a long time, astronomers have been trying to spot the dancers who are holding hands and spinning around each other (binary stars) while carrying a third partner (a planet) on their shoulders. These are called circumbinary planets.

Until now, finding these specific dancers has been like trying to spot a specific couple in a dark, crowded room using only a flashlight that flickers on and off. We've found a few dozen, mostly by watching them pass in front of their stars (transits) or by feeling the stars wobble slightly (radial velocity). But we don't really know how many of them exist, how heavy they are, or how far out they dance.

Enter Gaia: The Ultimate Dance Floor Camera

The European Space Agency's Gaia satellite is like a super-powered, high-definition camera that has been filming this dance floor for years. It doesn't just take pictures; it measures the exact position of every star with incredible precision. If a star is wobbling because of a planet, Gaia sees the star tracing a tiny loop in the sky.

This paper is a prediction of what Gaia will find when it releases its next batch of data (specifically versions DR4 and DR5, coming out around 2026). The authors, Thomas Baycroft and his team, built imaginary dance floors in a computer to see what would be found in each case, answering one big question: How many of these "triple-dance" planets will Gaia actually catch?

The Simulation: Building a Fake Universe

To make this prediction, the team didn't just guess. They constructed a virtual universe inside a computer:

The Cast: They started with a list of real stars Gaia has already seen (within 200 light-years of Earth).
The Couples: They randomly paired up these stars to create binary systems, using real statistics about how often stars have partners.
The Planets: They then "injected" planets into these systems. Since we don't know the true rules of how these planets form, they tested three main scenarios:
- Scenario 1: Maybe all these planets are huge gas giants (like Jupiter) versus medium-sized ones (like Saturn).
- Scenario 2: Maybe they huddle right next to the stars, or maybe they dance far away.

They ran this simulation 100 times for each scenario to see what Gaia would "see" under different conditions.

The Big Surprises

Here is what their "crystal ball" (the simulation) revealed:

1. The "Pile-Up" vs. The "Long Haul"
Previous studies thought these planets huddled very close to their stars (like a crowd at the center of the dance floor). But newer data suggests they might be spread out.

The Finding: While Gaia is indeed better at detecting planets that are far from their stars, the most important factor is the width of the binary star system itself. Since most binary stars are actually quite wide apart, Gaia will find the most planets if they are piled up close to these wide binaries.
The Analogy: Imagine trying to spot a runner on a track. If the stars are far apart (a wide track), even a planet orbiting relatively close to them creates a large, noticeable wobble in the star's position. Gaia is like a camera that can easily spot these large movements. Therefore, planets clustered near wide binaries are easier to detect than those orbiting tight binaries, regardless of the observation time.

2. The "Heavyweight" Bias
Gaia is like a scale that only tips if the weight is heavy enough.

The Finding: Gaia will mostly find massive planets (Super-Jupiters). It will struggle to see smaller, Saturn-sized planets unless they are very close to us.
The Result: Depending on which scenario is true, Gaia could find anywhere from 40 to nearly 600 new circumbinary planets. Even the "worst-case" scenario (40 planets) would double or triple the number of known circumbinary planets we have today!

3. The Mystery of the "Ghost Planets"
There is a group of controversial claims where astronomers think they see planets orbiting "dead" stars (white dwarfs) that used to be in a violent fight (a common-envelope phase). These claims are based on the stars' light flickering in a weird way.

The Finding: Many of these "ghost planets" are too far away or too small for Gaia to see. However, Gaia will be able to check about 3 to 11 of the most promising candidates.
The Impact: This is a huge deal. If Gaia sees them, they are real. If Gaia looks and sees nothing, those "ghosts" were just optical illusions caused by magnetic fields or other tricks. Gaia will act as the ultimate referee.

The Bottom Line

Think of this paper as a weather forecast for the future of astronomy.

The Forecast: We are going to get a massive influx of new data about planets orbiting two stars.
The Uncertainty: We don't know exactly how many yet because we don't know the "rules" of how these planets form.
The Excitement: No matter which scenario is right, Gaia is going to revolutionize our understanding. It will tell us if these planets are common or rare, if they are heavy or light, and whether they survive the violent deaths of their host stars.

In short, Gaia is about to turn the lights on in that dark dance floor, and we are going to finally see the whole dance troupe.

1. Problem Statement

The detection of circumbinary planets (CBPs) has historically been dominated by the transit method (e.g., Kepler), which suffers from strong geometric biases favoring short-period planets near the stability limit of the binary system. Radial velocity (RV) surveys have found fewer CBPs near this limit, instead detecting candidates at longer periods, creating a discrepancy in our understanding of the CBP population. Furthermore, the mass distribution of CBPs is poorly constrained, with many candidates (particularly those orbiting post-common-envelope binaries) being controversial due to reliance on eclipse timing variations (ETV) which can be mimicked by stellar activity.

The Gaia mission offers a new astrometric approach to detect CBPs, but previous yield estimates (e.g., Sahlmann et al. 2015) relied on outdated assumptions about the CBP population (specifically that all planets are "piled up" at $6 \times P_{bin}$ and follow the mass distribution of single-star planets). This paper aims to:

Re-evaluate Gaia's expected yield of CBPs using updated population statistics and sensitivity models.
Quantify how variations in assumed mass and orbital period distributions affect detection rates.
Assess Gaia's ability to validate or refute specific known CBP systems, particularly those orbiting post-common-envelope binaries.

2. Methodology

Synthetic Population Generation

The authors constructed a synthetic population of binary stars within 200 pc using Gaia DR3 sources.

Binary Injection: Using the Raghavan et al. (2010) multiplicity fractions and distributions, they injected binary companions into single-star Gaia sources. The binary fraction was set to ~40% (intermediate between "binarity" and "multiplicity" fractions) to account for hierarchical systems.
Parameters: Binary periods ( $P_{bin}$ ), mass ratios ( $q$ ), and eccentricities ( $e$ ) were drawn from distributions consistent with Raghavan et al. (2010), with specific cuts applied (e.g., $P_{bin} > 3$ days, removal of brown dwarf companions unless the primary is an M-dwarf).
Planet Injection: Gas giants and low-mass brown dwarfs ( $M \le 25 M_{Jup}$ $M \leq 25 M_{J u p}$ ) were injected into 10% of the binaries. The study tested 7 distinct injection setups to explore the parameter space:
- Mass Distributions: Combinations of a Gaussian distribution (peaking at Saturn mass) and a Log-Uniform distribution (extending to 25 $M_{Jup}$ ).
- Orbital Period Distributions: Defined in terms of scaled semi-major axis ( $a_{sc} = a_{pl} / r_{stab}$ ). Distributions included Log-Uniform ( $LU$ ), Uniform ( $U$ ), and fixed values ( $a_{sc}=1.2$ , representing the "pile-up" scenario).
- Fiducial Case (Setup 1): 40% Saturn-mass planets, 60% Log-Uniform mass distribution, and Log-Uniform period distribution.

Detectability Criteria

A planet was considered detectable if:

Astrometric SNR > 20: Calculated using the photocenter semi-major axis ( $\alpha$ ), number of observations ( $N_{obs}$ ), and astrometric uncertainty ( $\sigma$ ). The uncertainty model was updated based on Gaia DR3 performance (Holl et al. 2023a).
Orbital Coverage: The planet's orbital period ( $P_{pl}$ $P_{pl}$ ) must be less than the Gaia mission duration ( $T_{span}$ $T_{s p an}$ ).
- DR4: $T_{span} \approx 2000$ days.
- DR5: $T_{span} \approx 3800$ days.
Binary Detectability: The host binary itself must be astrometrically detectable (SNR > 20), except for known systems where this constraint is relaxed.

Analysis of Known Systems

The authors applied these criteria to:

Main-sequence binaries: Known transiting and RV-detected CBPs.
Post-common-envelope (PCE) binaries: A catalog of 32 claimed CBP candidates based on ETVs, calculating expected Gaia SNR for DR4 and DR5.

3. Key Results

Yield Predictions

Total Yield: Under conservative assumptions, Gaia is expected to detect tens to hundreds of CBPs.
- DR4: ~46 to 222 detections (depending on the injection setup).
- DR5: ~127 to 564 detections.
- The fiducial setup (Setup 1) predicts 100 detections in DR4 and 276 in DR5.
Comparison to Previous Estimates: These numbers are lower than the ~516 predicted by Sahlmann et al. (2015). The reduction is attributed to:
1. A more realistic mass distribution (fewer high-mass super-Jupiters).
2. A broader orbital period distribution (planets are not all clustered at the stability limit).
3. A more conservative binary population model.

Sensitivity to Population Assumptions

The study demonstrates that yield is highly sensitive to the assumed population properties:

Mass: Increasing the fraction of massive planets ( $>10 M_{Jup}$ ) significantly boosts the yield.
Orbital Period: Contrary to intuition, injecting planets closer to the binary stability limit (e.g., $a_{sc}=1.2$ ) yields fewer detections for short-period binaries but more for long-period binaries. This is because the underlying binary population has a peak frequency at longer periods (~100 days), and Gaia's astrometric sensitivity favors longer periods. Consequently, Gaia is biased toward detecting planets orbiting binaries with $P_{bin} > 50$ days, a regime largely unexplored by transit/RV surveys.

Known Systems Analysis

Main-Sequence Binaries: Gaia will not detect known transiting CBPs (e.g., Kepler-16b) due to their small astrometric signals and distance.
RV Candidates: Gaia DR4/5 will likely resolve the dichotomy in HD 202206 (distinguishing between a circumbinary brown dwarf and a two-planet system) and detect the massive companion in BEBOP-4 ( $\approx 20 M_{Jup}$ ).
Post-Common-Envelope Binaries:
- Of 32 claimed CBP candidates orbiting PCE binaries, Gaia DR5 is sensitive to 3 under strict criteria (SNR > 20, full orbit coverage) and 11 under relaxed criteria (SNR > 10, >50% coverage).
- The most promising candidates for confirmation are QS Vir, V470 Cam, and V893 Sco.
- Gaia will be able to test the reliability of ETV detections for PCE binaries, potentially confirming or refuting the existence of these massive companions.

4. Significance and Conclusions

Population Constraints: Gaia will provide the first large-scale, unbiased census of circumbinary planets, specifically constraining the mass and orbital period distributions for binaries with periods $P_{bin} > 50$ days. This will clarify whether the "pile-up" observed by Kepler is a universal feature or a selection effect.
Resolution of Controversies: The mission has the potential to definitively settle the debate regarding the existence of planets around post-common-envelope binaries, a population currently dominated by controversial ETV claims.
Methodological Impact: The paper highlights that yield predictions are highly dependent on the assumed underlying population. The wide variance in results across the 7 injection setups suggests that Gaia's actual detections will be a powerful tool for inferring the true CBP population properties rather than just confirming a pre-existing model.
Conservative Estimates: The authors emphasize that their predictions are lower bounds. Relaxing SNR thresholds or extending the distance limit would likely increase the yield, particularly for high-mass planets.

In summary, while Gaia may detect fewer CBPs than early optimistic estimates suggested, the quality and diversity of the sample will fundamentally alter our understanding of planet formation and survival in binary systems.