Bayesian hierarchical inference of asteroseismic inclination angles
Pith reviewed 2026-05-25 10:33 UTC · model grok-4.3
The pith
A Bayesian hierarchical model extracts stellar inclination angles from red giant asteroseismic data.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
The Bayesian hierarchical scheme provides a means to both accurately and robustly extract inclination angles from red giant stars by treating asteroseismic observables as the data source for per-star constraints while allowing joint inference across the sample.
What carries the argument
The Bayesian hierarchical scheme that pools information across stars while allowing individual inferences from asteroseismic mode amplitudes and frequencies.
If this is right
- Inclination angles for individual red giants can be recovered more robustly than with non-hierarchical analyses.
- Population distributions of inclinations become inferable once a selection function is included.
- Geometric properties of exoplanetary and binary systems gain better constraints from the derived angles.
- The method recovers an isotropic distribution when applied to artificial data generated under that assumption.
Where Pith is reading between the lines
- The same hierarchical structure could be adapted to infer other asteroseismic parameters such as rotation periods across a sample.
- Without a selection function, population inferences would systematically favor stars whose data are easier to observe.
- Combining the inclination results with spectroscopic or photometric rotation measurements could test consistency between independent observables.
Load-bearing premise
The method assumes that asteroseismic mode amplitudes and frequencies contain enough information to constrain inclination for each star individually and that a selection function can correct population-level biases when extending the model.
What would settle it
Applying the method to red giants with independent inclination measurements from transits or eclipsing binaries and finding large systematic mismatches with the hierarchical results.
read the original abstract
The stellar inclination angle-the angle between the rotation axis of a star and our line of sight-provides valuable information in many different areas, from the characterisation of the geometry of exoplanetary and eclipsing binary systems, to the formation and evolution of those systems. We propose a method based on asteroseismology and a Bayesian hierarchical scheme for extracting the inclination angle of a single star. This hierarchical method therefore provides a means to both accurately and robustly extract inclination angles from red giant stars. We successfully apply this technique to an artificial dataset with an underlying isotropic inclination angle distribution to verify the method. We also apply this technique to 123 red giant stars observed with $\textit{Kepler}$. We also show the need for a selection function to account for possible population-level biases, that are not present in individual star-by-star cases, in order to extend the hierarchical method towards inferring underlying population inclination angle distributions.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proposes a Bayesian hierarchical inference framework to extract stellar inclination angles from asteroseismic mode amplitudes and frequencies for individual red giant stars. It reports successful recovery of an isotropic distribution on simulated data, applies the method to 123 Kepler red giants, and notes that a selection function is required to extend the approach to population-level inferences without bias.
Significance. If the per-star posteriors are demonstrably data-driven rather than prior-dominated, the hierarchical scheme could provide a robust route to inclination angles for red giants, supporting studies of exoplanet obliquities and binary geometries. The framework's ability to share information across stars is a potential strength for noisy asteroseismic data.
major comments (2)
- [Abstract] Abstract: The artificial-dataset test is reported to recover the input isotropic population distribution, but the description provides no quantitative metrics (e.g., bias, scatter, or coverage of individual-star posterior medians versus true values) and does not compare hierarchical versus non-hierarchical single-star fits. This leaves open whether the reported 'accurate' single-star extractions are driven by the per-star likelihoods or by the shared population prior.
- [Abstract] Abstract: The central claim that the method 'accurately and robustly extract[s] inclination angles from red giant stars' on a per-star basis rests on the assumption that asteroseismic data constrain inclination independently of the hierarchical component. The reported verification tests only population-level recovery; it does not establish that individual posteriors remain data-dominated when mode-amplitude likelihoods are weak, as is common for red-giant mixed modes.
minor comments (1)
- [Abstract] Abstract: The sentence 'We also show the need for a selection function...' is repeated in slightly different form; a single clear statement of the selection-function requirement would improve readability.
Simulated Author's Rebuttal
We thank the referee for their constructive comments on our manuscript. We address the major comments point by point below and indicate the revisions we will incorporate.
read point-by-point responses
-
Referee: [Abstract] Abstract: The artificial-dataset test is reported to recover the input isotropic population distribution, but the description provides no quantitative metrics (e.g., bias, scatter, or coverage of individual-star posterior medians versus true values) and does not compare hierarchical versus non-hierarchical single-star fits. This leaves open whether the reported 'accurate' single-star extractions are driven by the per-star likelihoods or by the shared population prior.
Authors: We agree that the abstract lacks quantitative metrics for the recovery of individual inclinations and does not present a comparison to non-hierarchical fits. In the revised manuscript we will augment the abstract with summary statistics (median bias, rms scatter, and credible-interval coverage for the recovered individual inclinations relative to the known true values) and will add a brief statement noting that the hierarchical model yields tighter and better-calibrated posteriors than independent single-star analyses, particularly for stars with lower signal-to-noise modes. revision: yes
-
Referee: [Abstract] Abstract: The central claim that the method 'accurately and robustly extract[s] inclination angles from red giant stars' on a per-star basis rests on the assumption that asteroseismic data constrain inclination independently of the hierarchical component. The reported verification tests only population-level recovery; it does not establish that individual posteriors remain data-dominated when mode-amplitude likelihoods are weak, as is common for red-giant mixed modes.
Authors: The hierarchical construction is explicitly designed so that the population distribution acts as a shared prior while the per-star likelihood remains the dominant term when the data are informative. We acknowledge, however, that the validation presented in the current manuscript emphasises population-level recovery. In revision we will include an explicit demonstration that individual posteriors are data-driven: we will report the Kullback-Leibler divergence between the hierarchical posterior and the likelihood-only posterior for each star, and we will show that for the majority of the simulated and real targets the data contribution exceeds the prior contribution even in the presence of mixed-mode amplitude uncertainties. revision: yes
Circularity Check
No circularity; new hierarchical Bayesian framework applied to data with standard validation
full rationale
The paper introduces a Bayesian hierarchical scheme to infer stellar inclination angles from asteroseismic data on a per-star basis, then extends it to populations. The artificial dataset test recovers an input isotropic distribution at the population level, which is an independent verification step rather than a reduction of the per-star posteriors to a fitted parameter or self-definition. No equations or claims in the provided text reduce the central result to its inputs by construction, and no load-bearing self-citations or uniqueness theorems from the same authors are invoked. The derivation chain remains self-contained against external data.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.