APOKASC-3: Gaia-Anchored Evolved Star Catalog

• APOKASC-3 is a comprehensive catalog integrating APOGEE spectroscopy, Kepler asteroseismology, and Gaia parallaxes to derive precise stellar parameters for evolved stars.
• It employs ten independent asteroseismic pipelines and Gaia-derived luminosity scales to calibrate masses, radii, ages, and evolutionary states with high fidelity.
• The catalog serves as a benchmark for Galactic archaeology and stellar evolution, offering robust age scales and evolutionary labels for thousands of stars.

Searching arXiv for APOKASC-3 and closely related catalog papers. {"query":"APOKASC-3 Third Joint Spectroscopic and Asteroseismic catalog for Evolved Stars in the Kepler Fields arXiv APOKASC-3 related", "max_results": 10} APOKASC-3 is the third joint spectroscopic and asteroseismic catalog for evolved stars in the Kepler fields. It combines APOGEE high-resolution infrared spectroscopy, Kepler asteroseismology, and Gaia DR3 parallaxes, and it delivers calibrated evolutionary states, asteroseismic surface gravities, masses, radii, ages, and the spectroscopic and asteroseismic measurements used to derive them for evolved stars in the original Kepler field (Pinsonneault et al., 2024). In the catalog’s own framing, it is anchored to a fundamental radius scale from Gaia luminosities and spectroscopic effective temperatures rather than relying on uncalibrated scaling relations, and it is intended both as a precision stellar-parameter resource and as a Galactic-archaeology benchmark (Pinsonneault et al., 2024).

1. Origins within the APOKASC program

APOKASC originated as a joint effort between the Kepler Asteroseismic Science Consortium and the SDSS APOGEE spectroscopic survey, designed to combine seismic observables such as $\Delta\nu$ and $\nu_{\max}$ with APOGEE spectroscopy for stars observed by both surveys (Johnson, 2014). The first APOKASC catalog focused on about 1900 red giants and established the basic spectro-seismic architecture: APOGEE effective temperatures and metallicities, Kepler global oscillation parameters, and grid-based inference of $\log g$, mass, and radius (Johnson, 2014).

A major intermediate step toward APOKASC-3 was the consolidation of evolutionary-state classifications for APOKASC red giants. A consensus analysis of 6661 stars produced evolutionary-state labels for 6197 objects, using multiple seismic methods and explicitly preparing those labels for later grid-based mass and age inference (Elsworth et al., 2019). That work is important because APOKASC-3 treats RGB, red clump, secondary clump, and more luminous giants differently whenever the seismic information justifies it (Elsworth et al., 2019).

Against that lineage, APOKASC-3 is the large, Gaia-anchored culmination of the APOGEE–Kepler program for evolved stars. Its complete sample contains 15,808 evolved stars with APOGEE spectroscopic parameters and Kepler asteroseismology, and it provides full derived quantities for 12,418 stars while also retaining more limited seismic information for 1,624 additional objects outside the primary calibration domain or with lower-quality data (Pinsonneault et al., 2024).

2. Observational basis and inference architecture

The catalog is built from three observational pillars. APOGEE contributes calibrated $T_{\rm eff}$, metallicity, $\log g$, $\alpha$ abundances, and element ratios such as [C/Fe] and [N/Fe]; Kepler contributes long-cadence time-series photometry and the global asteroseismic observables $\nu_{\max}$ and $\Delta\nu$; Gaia DR3 contributes parallaxes and photometric information that are converted into luminosities and hence fundamental radii (Pinsonneault et al., 2024).

APOKASC-3 uses ten independent asteroseismic analysis techniques. Seven “core” pipelines return both $\nu_{\max}$ and $\Delta\nu$, and three auxiliary pipelines measure $\nu_{\max}$0 only; the pipeline outputs are then filtered, cross-compared, and combined (Pinsonneault et al., 2024). The catalog explicitly groups stars into Gold, Silver, Detection, and Non-Detection cohorts, with the Gold sample defined by the largest number of mutually consistent pipeline detections and therefore providing the highest-precision global parameters (Pinsonneault et al., 2024).

The central scaling framework is written with explicit correction factors:

$\nu_{\max}$1

$\nu_{\max}$2

In this formulation, $\nu_{\max}$3 is model-based and maps the observed large separation to mean density, whereas $\nu_{\max}$4 is empirically calibrated so that seismic radii match Gaia-based radii (Pinsonneault et al., 2024).

The Gaia anchoring is central. APOKASC-3 derives a reference radius scale from luminosity and spectroscopic $\nu_{\max}$5 and then calibrates the seismic radius scale to that reference. This suggests a deliberate shift from earlier APOKASC practice, in which cluster benchmarks played the dominant external role, toward a global radius system defined by Gaia parallaxes and photometry (Pinsonneault et al., 2024).

3. Calibration strategy, model dependence, and recommended domain

APOKASC-3 does not treat the scaling relations as exact. It computes alternative $\nu_{\max}$6 corrections using multiple stellar-model and weighting choices, including Garstec+Mosser, Garstec+White, and Sharma+White variants, specifically to expose model dependence in the mapping between $\nu_{\max}$7 and mean density (Pinsonneault et al., 2024). That design choice is itself one of the catalog’s methodological claims: multiple methods and explicit calibration are preferred to a single uncorrected seismic scale (Pinsonneault et al., 2024).

Within the calibrated regime, the catalog reports that scaling relations are precise and accurate on the lower RGB and in the red clump. Outside that regime, the behavior changes qualitatively. The catalog states that the relations become more model dependent for more luminous giants and break down at the tip of the RGB (Pinsonneault et al., 2024). For very low $\nu_{\max}$8 stars, APOKASC-3 therefore retains seismic observables and seismic $\nu_{\max}$9 but does not recommend the same level of trust in masses, radii, or ages (Pinsonneault et al., 2024).

A later reassessment of luminous RGB stars makes this limitation more explicit. In that study, an alternative interpretation of the APOKASC-3 correction scheme reduced fractional discrepancies in median masses and ages of lower RGB and upper RGB within the $\log g$0-rich population from $\log g$1 to $\log g$2 and from $\log g$3 to $\log g$4, respectively, indicating that the luminous-giant regime remained an active calibration problem after the main catalog release (Cao et al., 8 Aug 2025). A plausible implication is that APOKASC-3 should be used most conservatively where its own calibration claims are strongest: lower RGB and red-clump stars.

4. Catalog contents, cohorts, and precision

The catalog provides evolutionary state, asteroseismic surface gravity, mass, radius, age, and the underlying spectroscopic and asteroseismic measurements for 12,418 stars (Pinsonneault et al., 2024). It also includes 10,036 exceptionally precise measurements, with median fractional uncertainties in $\log g$5, $\log g$6, mass, radius, and age of $\log g$7, $\log g$8, $\log g$9, $T_{\rm eff}$0, and $T_{\rm eff}$1, respectively (Pinsonneault et al., 2024).

The evolved-star sample is not homogeneous in information content. APOKASC-3 distinguishes a Gold sample with the most robust multi-pipeline seismic measurements, a Silver sample with fewer but still usable detections, and additional stars with only partial or lower-quality seismic information (Pinsonneault et al., 2024). In the evolutionary-state domain, the detailed description reports 11,371 stars with robust seismic states, including 4,755 red-clump stars and 6,616 RGB/AGB stars (Pinsonneault et al., 2024).

The parameter inventory extends beyond the headline quantities. APOKASC-3 records spectroscopic inputs, seismic inputs, Gaia-based radii, alternative model-based corrections, evolutionary-state flags, and age estimates conditioned on evolutionary state; for luminous shell-burning stars, it provides asymptotic RGB and GB/AGB ages rather than a single universally preferred age (Pinsonneault et al., 2024). This reflects the catalog’s attempt to preserve provenance and model dependence rather than suppress it.

5. Astrophysical results and downstream scientific uses

One of APOKASC-3’s principal Galactic-archaeology results is a calibrated age scale for the chemical thick disk. Using lower RGB stars, the catalog finds a median age for the chemical thick disk of $T_{\rm eff}$2 Gyr with an age dispersion of 1.1 Gyr, and it also reports a sharp upper age boundary in the chemical thin disk (Pinsonneault et al., 2024). Earlier APOKASC-based Bayesian chemical-evolution modeling had already inferred a significant delay between two major gas-accretion episodes, with values in the range $T_{\rm eff}$3–$T_{\rm eff}$4 Gyr, and APOKASC-3 provides a more mature age platform for that class of inference (Spitoni et al., 2019).

The catalog has also become a reference system for external asteroseismic surveys. Ages in “The APO-K2 Catalog. II” agree with APOKASC-3 asteroseismic ages to within $T_{\rm eff}$5, which positions APOKASC-3 as a calibration and comparison scale for red-giant age work beyond the original Kepler field (Warfield et al., 2024).

APOKASC-3 has been used to revisit chemically anomalous populations. A study of young $T_{\rm eff}$6-rich stars from the third version of APOKASC found that stars with $T_{\rm eff}$7 were all single, whereas stars with $T_{\rm eff}$8 could be either single or binary, and concluded that many such stars do not follow the APOKASC stars in their [C/N], [C/Fe], and [N/Fe] trends with mass, favoring mass transfer for most of them (Jofre et al., 2022). This suggests that APOKASC-3 masses are astrophysically informative even when they no longer trace single-star ages in a straightforward way.

The catalog has also functioned as a calibration set outside Galactic archaeology. More than 2,000 RGB stars from APOKASC-3 were used to calibrate surface-brightness–color relations with asteroseismic radii and Gaia distances, and over the range $T_{\rm eff}$9–3 a change of 1 dex in [Fe/H] was found to modify the predicted angular diameter by less than 1% (Xiong et al., 22 Jun 2026). In that application, APOKASC-3 served less as an age catalog than as a homogeneous radius-and-metallicity standard.

6. Known systematics, contamination channels, and interpretation caveats

APOKASC-3 is precise, but it is not free of astrophysical contamination and spectroscopic systematics. A study of rapidly rotating red giants in APOKASC-3 identified 15,220 red giants in the control sample and found that $\log g$0 of them are rapid rotators at $\log g$1; it also identified a significant bias in reported metallicity for spectroscopically anomalous candidates, with median offsets of 0.37 dex in [M/H] relative to a control sample (Patton et al., 2023). The same work inferred a binary fraction of $\log g$2 for the full DR16 rapid-rotator candidate sample and argued that most rapid rotators are current or former binaries (Patton et al., 2023).

A related caveat concerns abundance-based age proxies. Work on abnormal carbon-to-nitrogen ratios in the APOKASC overlap sample found that at least 10% of red clump stars and approximately 10% of red giant branch stars deviate from the standard relationship between [C/N] and mass, with binary interactions responsible for the majority of these outliers (Bufanda et al., 2023). That result matters directly for APOKASC-3 interpretation because [C/N], [C/Fe], and [N/Fe] are part of the APOGEE information environment used in evolved-star inference.

The APOKASC lineage has also long documented binary-related seismic peculiarities. In the earlier APOKASC sample, only p-dominated mixed modes were found in about 4% of field red giants but in about 50% of red giants in detached eclipsing binaries, leading to a tentative binary-fraction estimate of about 8% for the APOKASC sample (Themeßl et al., 2017). This suggests that some apparently anomalous APOKASC-3 stars may be structurally non-standard rather than merely noisy.

Taken together, these results define the proper interpretive stance toward APOKASC-3. Within its calibrated regime, it is a high-precision Gaia-anchored spectro-seismic catalog. Outside that regime—or in populations affected by binarity, rapid rotation, mass transfer, abnormal [C/N], or luminous-giant scaling breakdown—its measurements remain valuable, but they must be read through the catalog’s own quality flags, evolutionary-state labels, and model-dependent alternatives rather than as universally interchangeable stellar ages or masses (Pinsonneault et al., 2024).