juliensimon/gaia-dr3-spectroscopic-binaries

---
license: cc-by-4.0
pretty_name: "Gaia DR3 Spectroscopic Binary Stars"
language:
- en
description: "Orbital solutions for 185,957 spectroscopic binary stars (SB1 + SB2) from ESA Gaia DR3, with periods, eccentricities, and radial velocity amplitudes."
task_categories:
- tabular-classification
tags:
- space
- open-data
- tabular-data
- parquet
- gaia
- esa
- binary-stars
- spectroscopic
- stellar
- astronomy
size_categories:
- 100K<n<1M
configs:
- config_name: default
data_files:
- split: train
path: data/gaia_dr3_spectroscopic_binaries.parquet
default: true
---

# Gaia DR3 Spectroscopic Binary Stars

<div align="center">
<img src="banner.jpg" alt="A youthful globular star cluster observed by the Hubble Space Telescope" width="400">
<p><em>Credit: NASA/ESA/Hubble</em></p>
</div>

*Part of the [Astronomy Datasets](https://huggingface.co/collections/juliensimon/astronomy-datasets-69c24caf2f17e36128946743) collection on Hugging Face.*

![Updated](https://img.shields.io/badge/dynamic/json?url=https://raw.githubusercontent.com/juliensimon/space-datasets/main/status.json&query=$["gaia-sb"]&label=updated&color=brightgreen)

Orbital solutions for **185,957** spectroscopic binary stars from the ESA Gaia Data
Release 3 non-single stars (NSS) pipeline -- **181,327** single-lined (SB1) and
**4,630** double-lined (SB2) systems with full Keplerian orbital parameters.

## Dataset description

Spectroscopic binaries are star systems where the binary nature is revealed not by eclipses
or visual separation, but by periodic Doppler shifts in the stellar spectral lines as the
stars orbit their common center of mass. In a **single-lined spectroscopic binary (SB1)**,
only one star's spectrum is visible -- the companion is too faint or its lines are blended
-- and the radial velocity curve of the primary yields the orbital period, eccentricity,
and the mass function f(m), which places a lower limit on the companion mass. In a
**double-lined spectroscopic binary (SB2)**, both stars contribute detectable spectral
lines, and the two radial velocity curves provide the mass ratio directly, constraining
individual masses when combined with inclination estimates.

The orbital elements in this catalog -- period, eccentricity, argument of periastron,
velocity semi-amplitudes (K1, K2), and systemic velocity -- are the fundamental observables
for binary star physics. They enable measurements of stellar masses (the single most
important parameter governing stellar evolution), tests of tidal circularization theory
(short-period binaries are expected to have circular orbits due to tidal dissipation),
and identification of compact companions such as white dwarfs, neutron stars, and stellar-mass
black holes lurking in otherwise unremarkable systems.

Gaia's contribution is transformative. The Radial Velocity Spectrometer (RVS) aboard Gaia
has obtained multi-epoch radial velocities for millions of stars brighter than about G ~ 12,
and the non-single stars pipeline has fitted Keplerian orbits to those showing statistically
significant radial velocity variability. The resulting catalog of over 185,957
spectroscopic binary orbits is an order of magnitude larger than any previous compilation --
the Ninth Catalogue of Spectroscopic Binary Orbits (SB9) contained fewer than 4,000 systems
accumulated over a century of ground-based observations. This leap in sample size enables
population-level studies of binary orbital properties (period distribution, eccentricity-period
relation, mass-ratio distribution) across a wide range of spectral types, metallicities, and
Galactic environments that were previously inaccessible.

The period-eccentricity diagram of these systems traces the physics of tidal evolution: systems
with periods below roughly 10 days are expected to have been circularized by tidal friction,
while longer-period systems retain the eccentricity imprinted at formation. The transition
between these regimes is a sensitive probe of stellar interior structure and the efficiency
of dissipative processes. Similarly, the distribution of mass functions and mass ratios
encodes the initial mass function of binary companions and the physics of binary star
formation -- whether companions are drawn randomly from the IMF or preferentially paired
with similar-mass primaries.

## Schema

| Column | Type | Description |
|--------|------|-------------|
| `solution_id` | int | Solution id |
| `source_id` | string | Gaia DR3 unique source identifier |
| `ra_deg` | float64 | Right ascension ICRS (degrees) |
| `dec_deg` | float64 | Declination ICRS (degrees) |
| `period_days` | float64 | Orbital period (days) |
| `period_days_err` | float64 | Period days err |
| `t_periastron` | float64 | Time of periastron passage (BJD) |
| `t_periastron_err` | float64 | T periastron err |
| `ecc` | float64 | Orbital eccentricity |
| `eccentricity_err` | float64 | Eccentricity err |
| `center_of_mass_velocity` | float64 | Center-of-mass velocity (km/s) |
| `center_of_mass_velocity_err` | float64 | Center of mass velocity err |
| `semi_amplitude_k1` | float64 | Semi amplitude k1 |
| `semi_amplitude_k1_err` | float64 | Semi amplitude k1 err |
| `omega` | float64 | Omega |
| `omega_err` | float64 | Omega err |
| `nrvp` | int | Nrvp |
| `ng_rvp` | int | Ng rvp |
| `nrvs` | int | Nrvs |
| `ng_rvs` | int | Ng rvs |
| `bit_index` | int | Bit index for solution flags |
| `corvec0` | float64 | Corvec0 |
| `corvec1` | float64 | Corvec1 |
| `corvec2` | float64 | Corvec2 |
| `corvec3` | float64 | Corvec3 |
| `corvec4` | float64 | Corvec4 |
| `corvec5` | float64 | Corvec5 |
| `corvec6` | float64 | Corvec6 |
| `corvec7` | float64 | Corvec7 |
| `corvec8` | float64 | Corvec8 |
| `corvec9` | float64 | Corvec9 |
| `corvec10` | float64 | Corvec10 |
| `corvec11` | float64 | Corvec11 |
| `corvec12` | float64 | Corvec12 |
| `corvec13` | float64 | Corvec13 |
| `corvec14` | float64 | Corvec14 |
| `obj_func` | float64 | Objective function value |
| `goodness` | float64 | Goodness |
| `eff` | float64 | Eff |
| `signi` | float64 | Signi |
| `flags` | int | Flags |
| `conf_persp` | float64 | Conf persp |
| `solution_type` | string | NSS solution type (SB1 = single-lined, SB2 = double-lined) |

## Quick stats

- **185,957** spectroscopic binary orbital solutions
- **181,327** single-lined (SB1) systems
- **4,630** double-lined (SB2) systems
- Period range: **0.2500** to **1451.96** days (median 118.5825)
- Median eccentricity: **0.193** (24,835 near-circular with e < 0.05)
- **43** columns with orbital, photometric, and astrometric parameters

## Usage

```python
from datasets import load_dataset

ds = load_dataset("juliensimon/gaia-dr3-spectroscopic-binaries", split="train")
df = ds.to_pandas()

# SB1 vs SB2 breakdown
print(df["solution_type"].value_counts())

# Period-eccentricity diagram (tidal circularization)
import matplotlib.pyplot as plt
period_col = [c for c in df.columns if "period" in c][0]
ecc_col = [c for c in df.columns if "eccentric" in c or c == "ecc"][0]
valid = df.dropna(subset=[period_col, ecc_col])
plt.scatter(valid[period_col], valid[ecc_col], s=0.3, alpha=0.3)
plt.xscale("log")
plt.xlabel("Period (days)")
plt.ylabel("Eccentricity")
plt.title("Gaia DR3 Spectroscopic Binaries: Period vs Eccentricity")
plt.show()

# High-eccentricity systems
high_ecc = df[df[ecc_col] > 0.8]
print(f"{len(high_ecc):,} systems with eccentricity > 0.8")
```

## Data source

Gaia Collaboration, Arenou, F. et al. (2023), *Gaia Data Release 3: stellar multiplicity,
a teaser for the hidden treasure.* Astronomy & Astrophysics, 674, A34.
European Space Agency, accessed via [VizieR](https://vizier.cds.unistra.fr/) CDS Strasbourg
(I/357) and/or the [Gaia Archive](https://gea.esac.esa.int/archive/).

## Update schedule

Static dataset -- uploaded once from Gaia DR3. Will be updated when Gaia DR4 is released.

## Related datasets

- [wds-double-stars](https://huggingface.co/datasets/juliensimon/wds-double-stars) -- Washington Double Star Catalog
- [gaia-dr3-eclipsing-binaries](https://huggingface.co/datasets/juliensimon/gaia-dr3-eclipsing-binaries) -- Gaia DR3 Eclipsing Binaries
- [kepler-eclipsing-binaries](https://huggingface.co/datasets/juliensimon/kepler-eclipsing-binaries) -- Kepler Eclipsing Binary Catalog
- [xray-binary-catalog](https://huggingface.co/datasets/juliensimon/xray-binary-catalog) -- X-ray Binary Catalog

## Pipeline

Source code: [juliensimon/space-datasets](https://github.com/juliensimon/space-datasets)

## Support

If you find this dataset useful, please give it a ❤️ on the [dataset page](https://huggingface.co/datasets/juliensimon/gaia-dr3-spectroscopic-binaries) and share feedback in the Community tab! Also consider giving a ⭐️ to the [space-datasets](https://github.com/juliensimon/space-datasets) repo.

## Citation

```bibtex
@dataset{gaia_dr3_spectroscopic_binaries,
author = {Simon, Julien},
title = {Gaia DR3 Spectroscopic Binary Stars},
year = {2026},
publisher = {Hugging Face},
url = {https://huggingface.co/datasets/juliensimon/gaia-dr3-spectroscopic-binaries},
note = {Based on Gaia DR3 (Arenou et al. 2023, A&A 674, A34) via VizieR CDS Strasbourg}
}
```

## License

[CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/)