Identification of 30,000 White Dwarf-Main Sequence binaries candidates from Gaia DR3 BP/RP(XP) low-resolution spectra Facilities: Gaia
Jiadong Li, Yuan-Sen Ting, Hans-Walter Rix, Gregory M. Green, David W. Hogg, Juan-Juan Ren, Johanna Müller-Horn, Rhys Seeburger · 2025
Read Identification of 30,000 White Dwarf-Main Sequence binaries candidates from Gaia DR3 BP/RP(XP) low-resolution spectra Facilities: Gaia by Jiadong Li, Yuan-Sen Ting, Hans-Walter Rix, Gregory M. Green, David W. Hogg, Juan-Juan Ren, Johanna Müller-Horn, Rhys Seeburger (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-14494.txt.
Introduction
Jiadong Li and co-authors fish thirty thousand candidate double stars out of the Gaia catalog. Each pair joins a white dwarf, a dead star's dense core, to a living main-sequence star. Such pairs teach how binary stars age together and seed several kinds of cosmic explosions, yet they hide well when the bright partner drowns the dim one.
The team trained a neural network on low-resolution Gaia spectra and sorted pairs with a Gaussian process classifier. From more than ten million stars within a kiloparsec, about thirty thousand candidates surfaced. Spectral fitting confirmed some seventeen hundred with high confidence. Ultraviolet checks backed the list: most of the right candidates glowed with the excess that betrays a white dwarf.
Catalog record
| CREATOR | Jiadong Li, Yuan-Sen Ting, Hans-Walter Rix, Gregory M. Green, David W. Hogg, Juan-Juan Ren, Johanna Müller-Horn, Rhys Seeburger |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 70 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-14494
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Identification of 30,000 White Dwarf-Main Sequence binaries candidates from Gaia DR3 BP/RP(XP) low-resolution spectra Facilities: Gaia Jiadong Li, Yuan-Sen Ting, Hans-Walter Rix, Gregory M. Green, David W. Hogg, Juan-Juan Ren, Johanna Müller-Horn, Rhys Seeburger Abstract White dwarf-main sequence (WDMS) binary systems are essential probes for understanding binary stellar evolution and play a pivotal role in constraining theoretical models of various transient phenomena. In this study, we construct a catalog of WDMS binaries using Gaia DR3’s low-resolution BP/RP (XP) spectra. Our approach integrates a model-independent neural network for spectral modelling with Gaussian Process Classification to accurately identify WDMS binaries among over 10 million stars within 1 kpc. This study identify approximately 30,000 WDMS binary candidates, including 1,700 high-confidence systems confirmed through spectral fitting. Our technique is shown to be effective at detecting systems where the main-sequence star dominates the spectrum - cases that have historically challenged conventional methods. Validation using GALEX photometry reinforces the reliability of our classifications: 70% of candidates with an absolute magnitude M G > 7 M_{G}>7 exhibit UV excess, a characteristic signature of white dwarf companions. Our all-sky catalog of WDMS binaries expands the available dataset for studying binary evolution and white dwarf physics and sheds light on the formation of WDMS. PyTorch [ 39 ] , Astropy [ 3 ] , laspec [ 63 ] , Scipy [ 58 ] , scikit-learn [ 40 ] , TOPCAT [ 56 ] , I Introduction Binary star systems containing a white dwarf (WD) and a main sequence (MS) companion (WDMS binaries) are fundamental laboratories for studying stellar evolution in interacting systems. The evolution of WDMS binaries follows two distinct pathways, determined by their initial orbital separations [ 9 , 60 ] . In approximately 75% of the cases, the initial separations are sufficiently wide ( \gtrsim 10 AU) that the more massive star evolves essentially as a single star, eventually becoming a white dwarf without interaction with the companion [ 17 ] . These systems maintain orbital separations similar to their initial configurations. The remaining \sim 25% of systems experience a more dramatic evolution. As the more massive star expands into the giant phase, the binary undergoes a Common Envelope (CE) phase in which orbital energy and angular momentum losses lead to a dramatic decrease in orbital separation [ 29 ] , although recently putative PCEBs have been found in wide orbits [ 61 ] . This process results in the ejection of the envelope and the formation of a post-common envelope binary (PCEB). The PCEBs play an important role in a wide range of problems, e.g., constraining the effeciency of common evelope evolution [ 47 , 67 ] , These PCEBs are the progenitors of various exotic systems, including Type Ia supernovae (e.g., Han & Podsiadlowski 26 , Wang & Han 59 , Hernandez et al. 27 ) , double degenerate binaries, cataclysmic variables [ 37 , 55 ] , and low-mass X-ray binaries [ 66 ] . The identification and characterization of WDMS binaries have historically been challenging due to the vast difference in luminosities between components, particularly in systems with AFGK-type stars. White dwarfs are typically much fainter than their main sequence companions, making their spectral signatures difficult to detect in unresolved systems. Different techniques have been developed to identify these systems: spatially resolved wide binaries can be detected through common-proper motion pairs, while unresolved systems require spectroscopic analysis or the detection of eclipses. Using common-proper motion, Gaia [ 18 ] EDR3 identified approximately 16,000 wide WDMS systems within 1 kpc [ 15 ] . For unresolved systems where one of the components dominates the spectral energy distribution (estimated to be 90% of cases), identification becomes challenging [ 48 ] . Large spectroscopic surveys have made great advances in identifying WDMS binaries, with the Sloan Digital Sky Survey (SDSS; York et al. 62 ) identifying over 3,200 WDMS systems [ 46 ] and the Large Sky Area Multi-Object Fiber Spectroscopic Telescope (LAMOST; Deng et al. 10 ) survey discovering approximately 1,000 additional pairs [ 50 , 49 ] by spectra with resolution \sim 2000 . These samples have been explored to study CE evolution [ 66 ] as well as to test mass-radius relationships [ 38 ] . However, both SDSS and LAMOST samples are subject to selection effects and are particularly biased against systems where one component dominates the spectral energy distribution (SED) [ 45 ] . These limitations manifest themselves in two key aspects. First, the traditional identification methods, which rely on \chi^{2} -fitting, wavelet transforms, and human visual inspection, struggle to identify the WDMS that has a component dominated the SED. Recent studies have shown that such systems can comprise up to 91% of the total population of WDMS binaries [ 48 ] , making this a limitation in our understanding of these systems. Recent work combining UV and optical data has helped identify some previously missed hot WD systems [ 36 ] , many WDMS binaries likely remain undetected. Second, existing established approaches to identifying WDMS binaries have their own challenges. They either rely on previously identified SDSS/LAMOST systems as training sets, potentially inheriting the same selection biases [ 48 ] , or depend on synthetic spectral libraries to construct theoretical WDMS binary templates [ 13 ] . Although theoretical models have made progress for cool dwarf stars [ 2 ] , there remain some systematic differences between synthetic and observed spectra, especially for cool dwarf stars [ 32 , 41 ] , which can affect the accuracy of binary identification. These model limitations can lead to erroneous identification and characterization. The Gaia mission [ 18 ] opens a new window in the study of WDMS binaries [ 36 , 24 ] over a wide range of stellar parameters and distances [ 13 ] , with well-understood selection effects and homogeneous data quality. In this paper, we present a new method for detecting WDMS binaries using Gaia DR3 XP spectra. Our approach combines the interpretability of \chi^{2} fitting with machine learning techniques. We first develop data-driven spectral emulators for single main-sequence stars, white dwarfs, and WDMS binaries, following a similar iterative approach to eliminate binary contamination in the training sample as described by El-Badry et al. [16] . This ensures that our models are built from observations themselves, independent of theoretical spectral templates. We then employ a Gaussian Process Classifier (GPC) trained on these candidates to identify WDMS binaries across the full parameter space - particularly targeting systems where one component dominates the spectral energy distribution, which are challenging to detect through \chi^{2} fitting alone. Combining the high-quality homogeneous data provided by Gaia with this hybrid methodology, we aim to construct the most complete sample of WDMS binaries from spectrophotometric data from Gaia . The remainder of the paper is structured as follows: In Section II , we describe the Gaia DR3 data and our sample selection criteria. Section III outlines our spectral fitting methodology and the development of our single-star and binary models. In Section IV , we present our classification approach using our Gaussian Process classification techniques. Section V details the results of our WDMS binary detection, including a discussion of the reliability of our method. We summarize our