Asteroseismic Binaries Research

• Asteroseismic binaries are unresolved systems where both stars exhibit detectable solar-like oscillations, enabling combined seismic and binary analysis.
• Detection relies on high-cadence Kepler data and spectroscopic vetting to separate overlapping power spectra and mitigate contamination.
• Studies of these binaries constrain stellar masses, radii, and ages, offering unique tests of stellar evolution and binary interaction models.

Searching arXiv for recent and foundational papers on asteroseismic binaries. Search query: asteroseismic binaries Kepler unresolved red giant subgiant eclipsing binary Asteroseismic binaries (ABs) are systems in which two oscillating stars are observed in the same light curve, so the frequency or power-density spectrum contains two stellar oscillation signatures; in the Kepler solar-like and red-giant literature this usually denotes two stars with detectable solar-like oscillations in a single combined light curve, while some binary-pulsation literature uses the term more broadly for binaries discovered or characterized through pulsations of one or both components (Miglio et al., 2014, Espinoza-Rojas et al., 16 Sep 2025, Shibahashi et al., 2019). Their astrophysical value lies in the conjunction of binary constraints—common origin, common age, common initial composition, and in favorable systems dynamical masses and radii—with asteroseismic constraints on mass, radius, evolutionary state, internal structure, core properties, internal mixing, and overshooting (Murphy, 2018, Southworth, 2015, Murphy, 2024).

1. Definition, scope, and terminology

In the narrow sense now common in Kepler red-giant and solar-like work, an AB is an unresolved system in which both components show detectable solar-like oscillations in the combined light curve, producing two oscillation envelopes in the power spectrum; these envelopes may be well separated or may overlap strongly (Miglio et al., 2014, Mazzi et al., 28 Apr 2025). This definition is deliberately narrower than the broader category of binary-evolution products with detectable oscillations, which includes systems altered by mass transfer, stripping, mergers, or common-envelope evolution but not necessarily showing oscillations from both components (Mazzi et al., 28 Apr 2025).

The observational category does not by itself guarantee that the two oscillating sources are gravitationally bound. Recent Kepler–Gaia work explicitly treats ABs as systems in which two oscillating stars are observed in the same light curve, arising either from chance alignments or from gravitationally bound stars (Espinoza-Rojas et al., 16 Sep 2025). Asteroseismic analyses of seismically unresolved candidates likewise show that some observational ABs are chance alignments rather than physical binaries (Choi et al., 27 Apr 2026). When the pair is genuinely binary, however, the common age and initial composition become stringent priors for stellar modelling, and in wide systems the stars can usually be modeled with single-star evolution tracks because tidal interaction and mass transfer are negligible (Schimak et al., 19 Jan 2026).

A broader binary-seismology usage survives in work on coherent pulsators. In that framework, binaries can be discovered and characterized through phase modulation of a pulsating star, with orbital motion encoded in light-travel-time delays of the pulsation phase; this approach uncovered more than 340 non-eclipsing binaries with A- and F-type primaries and included 21 double-pulsator binaries, but it does not work well for stochastically excited solar-like oscillations (Shibahashi et al., 2019). This terminological divergence is substantial: in current red-giant and solar-like practice, “AB” usually implies two detectable oscillators in one light curve, whereas in broader pulsation-binary work it can denote binaries inferred from pulsations more generally (Shibahashi et al., 2019, Murphy, 2024).

2. Detection channels and observational vetting

The defining observational signature of a solar-like AB is a single Kepler light curve whose spectrum contains two solar-like oscillation power excesses (Miglio et al., 2014). Cadence is decisive. For cool main-sequence stars and subgiants, short cadence is required because their oscillation periods are minutes long; Kepler short cadence is about 58.85 s, whereas long cadence is about 29.4 min and is primarily suitable for red giants below the Nyquist frequency near $283\,\mu{\rm Hz}$ (Miglio et al., 2014). KIC 7107778 exemplifies the short-cadence regime: it was observed in Kepler short cadence over nearly two years, and only short-cadence data were usable because its oscillation power is centered near $\sim 550\,\mu{\rm Hz}$, well above the long-cadence Nyquist frequency (Li et al., 2018).

Population-level detection work formalized this signature with standard solar-like detectability calculations. In the original Kepler prediction study, a detection was flagged when the probability of detection exceeded 90%, with oscillation and granulation amplitudes corrected for flux dilution from the companion (Miglio et al., 2014). A later synthetic-population study adopted the same 90% threshold for both components over the full four-year Kepler time series and required the combined binary detectability to exceed 0.9 (Mazzi et al., 28 Apr 2025). This detectability formalism immediately explains why red-giant ABs dominate the expected sample: long-cadence data suffice for red giants, whereas solar-type ABs require scarce short-cadence coverage (Miglio et al., 2014).

Spectroscopic vetting is a parallel requirement because unresolved companions bias the atmospheric and photometric inputs used in seismic modeling. Ground-based high-resolution spectroscopy of Kepler seismic targets was used to determine $T_{\rm eff}$, $\log g$, $[\mathrm{Fe/H}]$, $v\sin i$, and $v_r$, while simultaneously identifying hidden SB1 and SB2 systems that could contaminate seismic interpretation (Molenda-Zakowicz et al., 2010, Molenda-Zakowicz et al., 2012). One program found three new SB2s and two suspected SB1s among 44 Kepler asteroseismic targets, and another identified six SB1s and two SB2s in a 95-star Kepler-field sample (Molenda-Zakowicz et al., 2010, Molenda-Zakowicz et al., 2012). The motivation was explicit: neglecting duplicity corrections can lead to an incorrect asteroseismic model because an unresolved secondary changes the observed flux, colors, luminosity, inferred temperature, and sometimes even mode identification (Molenda-Zakowicz et al., 2012).

Eclipsing and hierarchical systems require additional spatial vetting. In the Kepler red-giant/eclipsing-binary cross-match, eclipse modeling, target-pixel contamination checks, and asteroseismology were used jointly because catalog coincidence alone did not guarantee that the pulsating red giant and the eclipsing source were the same physical object (Gaulme et al., 2013). Target-pixel tests rejected several false associations by showing that the eclipse signal originated on different pixels from the red-giant light source (Gaulme et al., 2013). This established a general methodological point that later became central in unresolved-AB work: in crowded Kepler apertures, seismology without pixel-level source attribution is not sufficient (Gaulme et al., 2013, Choi et al., 27 Apr 2026).

3. Frequency-domain phenomenology: overlap, morphology, and hidden binaries

AB phenomenology is dominated by the degree to which the two oscillation spectra overlap. KIC 7107778 is the canonical subgiant case: a non-eclipsing, unresolved target whose two components both show solar-like oscillations in a single Kepler light curve, with completely overlapping oscillation power profiles (Li et al., 2018). The two stars were assigned by their mean large separations, $$\langle\Delta\nu\rangle_A = 31.83 \pm 0.02\,\mu{\rm Hz}$$ and $$\langle\Delta\nu\rangle_B = 34.55 \pm 0.01\,\mu{\rm Hz}$$, while $l=0$ and $\sim 550\,\mu{\rm Hz}$0 modes formed clean ridges in the echelle diagram and most $\sim 550\,\mu{\rm Hz}$1 modes were mixed modes perturbed by avoided crossings (Li et al., 2018). The system demonstrated that nearly identical subgiants can occupy almost the same frequency domain and still be separated seismically.

Synthetic unresolved red-giant ABs show that this is not an exceptional pathology but a generic regime. A set of 5,000 artificial asteroseismic binary systems constructed from Kepler red-giant KASOC light curves showed that the majority consist of two red-clump stars and that unresolved ABs generally exhibit increased Shannon entropy and decreased oscillation power compared to individual components (Choi et al., 2 Jun 2025). For similar-brightness systems, four PDS morphologies were defined: single star-like, aligned, partially aligned, and misaligned. Aligned systems typically have $\sim 550\,\mu{\rm Hz}$2 and $\sim 550\,\mu{\rm Hz}$3 frequency differences of less than about 10% of the mean $\sim 550\,\mu{\rm Hz}$4; partially aligned cases typically fall around 10% to 25%; misaligned cases generally exceed about 25% and produce highly complex oscillation patterns (Choi et al., 2 Jun 2025). This morphology-based description reframed unresolved AB detection as a problem of spectral organization rather than merely amplitude dilution.

Observed unresolved candidates confirm that blended oscillations can corrupt mode identification and stellar inference. In six Kepler red-giant candidates whose PDS morphologies changed across different light-curve extractions, oscillations from two stars produced inaccurate mode identification and biased global seismic parameters (Choi et al., 27 Apr 2026). The reported biases reached up to about 30% in $\sim 550\,\mu{\rm Hz}$5 and about 35% in $\sim 550\,\mu{\rm Hz}$6; inferred masses and radii differed by up to about 3 and 2 times, respectively, relative to the individual stars (Choi et al., 27 Apr 2026). In the most complex case, the coupling factor became unreliable and was often overestimated (Choi et al., 27 Apr 2026). A central consequence is that complex or low-power PDS morphologies are not uniquely single-star anomalies; unresolved AB superposition is a documented alternative explanation (Choi et al., 2 Jun 2025, Choi et al., 27 Apr 2026).

4. Benchmark systems and forward modelling

Several systems have become benchmarks because they expose different parts of the AB problem, from nearly twin subgiants to red-clump model failures.

System	Configuration	Seismic significance
KIC 7107778	non-eclipsing unresolved twin subgiants	first detected asteroseismic sub-giant binary; fully overlapping power profiles (Li et al., 2018)
HD 176465	binary with both components as solar-like pulsators	coeval ages, common initial composition, He II acoustic-glitch helium abundances (Gai et al., 2018)
16 Cyg A/B	solar-analog binary modeled independently	common age $\sim 550\,\mu{\rm Hz}$7 Gyr and common composition (Metcalfe et al., 2015)
KIC 10841730	wide RC+RGB binary, $\sim 550\,\mu{\rm Hz}$8 d	first study of an AB consisting of an RC star and an RGB star; exposes RC-model tensions (Schimak et al., 19 Jan 2026)

KIC 7107778 established the unresolved-subgiant template. Oscillation mode parameters were extracted with Bayesian inference and nested sampling using DIAMONDS, fitting a white-noise term, three Harvey-like granulation components, a Gaussian envelope, and 32 Lorentzians (Li et al., 2018). Stellar models were computed with MESA, oscillation frequencies with GYRE, and near-surface effects corrected with

$\sim 550\,\mu{\rm Hz}$9

The final grid-modelling results gave $T_{\rm eff}$0, $T_{\rm eff}$1, $T_{\rm eff}$2, $T_{\rm eff}$3, $T_{\rm eff}$4 Gyr, and $T_{\rm eff}$5 Gyr, with a mass difference of only about $T_{\rm eff}$6 (Li et al., 2018). The paper interpreted this as a widely separated twin binary whose simultaneous birth and evolution were confirmed seismically (Li et al., 2018).

HD 176465 showed how binary coevality can be combined with acoustic-glitch diagnostics. Using YREC models and individual low-degree frequencies, the two components were found to have consistent ages and the same initial helium abundance and metallicity within uncertainties, while the He II acoustic glitch yielded current envelope helium abundances $T_{\rm eff}$7 and $T_{\rm eff}$8 (Gai et al., 2018). The same analysis constrained the acoustic depths of the He II ionization zone and the base of the convection zone, and found that both the amplitude and acoustic depth of these glitch signatures are functions of mass (Gai et al., 2018). ABs therefore act ոչ միայն as coevality tests but also as laboratories for internal stratification and diffusion.

16 Cyg A and B occupy a somewhat different niche: they are a bright solar-analog binary used to test whether independent seismic modeling recovers a shared age and composition (Metcalfe et al., 2015). Modeling with AMP and the complete 30-month Kepler data set returned the same age, $T_{\rm eff}$9 Gyr, and common composition, $\log g$0 and $\log g$1, for the two stars (Metcalfe et al., 2015). Because the stars were fit independently, the agreement served as a stringent validation of the modelling framework rather than as a built-in prior (Metcalfe et al., 2015).

KIC 10841730 extended AB work into the core-helium-burning regime. It is a wide binary containing a red-clump star and an RGB star, modeled with individual $\log g$2 frequencies, $\log g$3 modes for $\log g$4, and a large MESA+GYRE grid spanning mass, metallicity, helium abundance, mixing length, overshooting, and mass loss (Schimak et al., 19 Jan 2026). The RGB star was modeled satisfactorily, but the RC component admitted two incompatible interpretations: one required an unexpected positive offset in the p-mode phase shift and yielded a binary-consistent solution with current masses $\log g$5 and $\log g$6 for the RC and RGB stars, whereas the other removed two questionable radial modes and implied a much younger RC star inconsistent with the companion age (Schimak et al., 19 Jan 2026). The system therefore showed that uncertainties in RC models affect not only $\log g$7 but also the p-mode spectrum (Schimak et al., 19 Jan 2026).

5. Eclipsing and spectroscopic calibrators

Eclipsing and spectroscopic binaries provide the dynamical benchmark against which seismic inferences can be tested. In the Kepler red-giant/eclipsing-binary program, simultaneous modeling of eclipse photometry and RV curves with JKTEBOP yielded absolute masses and radii with typical precisions of roughly 3.35% for red-giant masses and 1.04% for red-giant radii (Gaulme et al., 2016). Using 10 pulsating red giants in SB2 systems as the main comparison set, the study found that asteroseismic scaling relations systematically overestimate red-giant radii by $\log g$8 and masses by $\log g$9 when Mosser et al. corrections are applied; standard scalings gave even larger offsets (Gaulme et al., 2016). The consequence is direct: systematic mass overestimation implies systematic age underestimation in ensemble red-giant work (Gaulme et al., 2016).

The same eclipsing-binary sample also showed that oscillations can be suppressed in close red-giant binaries. Four non-oscillating candidates all lay among the shortest-period systems, showed strong activity and likely tidal synchronization, and had expected $[\mathrm{Fe/H}]$0 values of 27, 59, 61, and 74 $[\mathrm{Fe/H}]$1Hz, well below the Kepler long-cadence Nyquist frequency (Gaulme et al., 2016). The lack of oscillation detection was therefore not a cadence artifact but evidence that oscillations in close red-giant binaries can be suppressed enough to become undetectable (Gaulme et al., 2016). This placed tides and magnetic activity directly inside the seismic selection function.

Catalog-level RG/EB work established the size and complexity of the calibration sample. Cross-matching the Kepler red-giant catalog of 13,698 stars with the eclipsing-binary catalog of 2,616 systems produced 70 RG/EB candidates (Gaulme et al., 2013). After light-curve modeling, contamination checks, and plausibility tests, the final interpretation yielded 13 strong bona-fide pulsating RG/EB systems, 1 bona-fide non-eclipsing heartbeat binary, and 10 likely hierarchical triple systems, with an additional 5 non-pulsating long-period detached systems considered likely true RG/EBs (Gaulme et al., 2013). These systems supplied not just dynamical checks on seismic mass and radius, but also mixed-mode classifications, g-mode period spacings, and core-rotation measurements from roughly 30 to 385 days (Gaulme et al., 2013).

Ground-based spectroscopic programs complete this calibration chain by identifying which seismic targets are composite before any seismic modeling is attempted. Spectroscopic Kepler follow-up discovered multiple SB1 and SB2 systems, derived orbital elements for HIP 97321, and emphasized that the mass function in SB1s and orbital solutions in SB2s are essential because unresolved companions otherwise bias the inferred atmospheric parameters used for asteroseismic modeling (Molenda-Zakowicz et al., 2010, Molenda-Zakowicz et al., 2012). The binary role is therefore double-edged in the precise sense stated by the literature: binarity can contaminate the observed light and distort the stellar parameters used in seismic analysis, but eclipsing SB2 systems are among the strongest calibrators in stellar astrophysics (Molenda-Zakowicz et al., 2012).

6. Demographics, binary evolution, and current tensions

Population-synthesis studies originally suggested that ABs should be a substantial though highly specialized Kepler population. Simulations of the Kepler field with TRILEGAL and BiSEPS predicted about 200 or more detectable ABs, mostly pairs of He-core-burning red giants, with only a small number of solar-type systems because of limited short-cadence coverage (Miglio et al., 2014). The predicted yield depended strongly on the assumed initial mass-ratio distribution, and the detectable systems were expected to have mass ratios within roughly 10% of unity (Miglio et al., 2014). A key implication was that AB counts could act as a sensitive calibration of the debated mass-ratio distribution near $[\mathrm{Fe/H}]$2 (Miglio et al., 2014).

Later synthetic-population work added binary interaction and reduced the expected occurrence rate among detectable red giants. In DUETS, the fraction of red giants with detectable oscillations that are ABs ranged from about 0.46% for non-interacting binaries to about 0.06% in an interacting Eggleton-type simulation, and in one MDS17 realization only 227 of 41,537 detectable red giants were ABs (Mazzi et al., 28 Apr 2025). The AB population was dominated by double CHeB systems, CHeB+RGB systems, and double RGB systems, and 95% of ABs had initial mass ratios $[\mathrm{Fe/H}]$3 (Mazzi et al., 28 Apr 2025). The same study argued that even small mass-transfer events can prevent the detection of oscillations from both components and that double red-clump systems are not expected at separations $[\mathrm{Fe/H}]$4 because RGB expansion and Roche-lobe overflow disrupt the route to a surviving double-red-clump AB (Mazzi et al., 28 Apr 2025).

Observed Kepler samples have complicated this demographic picture. A Gaia-based analysis of 40 well-resolved ABs in the Kepler field found that most appear to be chance alignments rather than gravitationally bound binaries; only KIC 6501237 and KIC 10094545 emerged as promising wide-binary candidates, with probabilities of about 50% and about 20–25%, respectively, and 11 systems were likely spatially unresolved multiples based on Gaia multiplicity indicators (Espinoza-Rojas et al., 16 Sep 2025). This does not directly refute the earlier synthetic forecasts because the observed sample was biased toward seismically resolved systems, whereas the population models and unresolved-PDS work imply that many true ABs should be seismically unresolved and therefore harder to recognize (Espinoza-Rojas et al., 16 Sep 2025, Choi et al., 2 Jun 2025).

Binary evolution introduces a further tension between ABs as clean coeval laboratories and oscillating stars as products of past interaction. At least about 1% of Kepler red giants with detectable oscillations may have undergone significant mass accretion or loss, creating over-massive or under-massive stars that bias age–metallicity relations if interpreted as single-star products (Mazzi et al., 28 Apr 2025). In massive early-type accretors, binary mass transfer can leave distinctive g-mode signatures; the recent MESA+GYRE parameter study of a 4.0 $[\mathrm{Fe/H}]$5 donor plus 3.0 $[\mathrm{Fe/H}]$6 accretor showed that convective boundary mixing dominates rejuvenation and that post-accretion thermal relaxation is crucial for producing the final seismic imprint, including multi-component Fourier transforms of period-spacing patterns (Henneco et al., 11 Jun 2026). ABs are therefore both precision benchmarks and diagnostics of binary interaction, but the latter role requires that binary history not be ignored.

The current field is defined by three linked open problems. First, observational AB samples remain incomplete because unresolved systems can mimic single stars and because many resolved Kepler pairs are not physically bound (Choi et al., 2 Jun 2025, Espinoza-Rojas et al., 16 Sep 2025, Choi et al., 27 Apr 2026). Second, the best benchmark systems are now exposing deficiencies in stellar models, especially for red-clump stars, where binary consistency tests reveal that standard RC models can fail in both mixed-mode and p-mode diagnostics (Schimak et al., 19 Jan 2026). Third, AB occurrence rates and morphologies are sensitive to binary initial conditions, mass-ratio distributions, mass transfer, and tidal evolution, so comparisons between predicted and observed AB populations remain a direct route to constraining binary-star physics (Miglio et al., 2014, Mazzi et al., 28 Apr 2025).