Asteroseismology of SPB stars: a comparison of forward asteroseismic modelling results from Kepler and TESS

TL;DR

This study assesses whether short, intermittent TESS data can reproduce forward asteroseismic modelling results previously obtained from 4-year Kepler light curves for SPB stars. By re analysing Kepler data with a common extraction method and extracting frequencies from 1–3 sector TESS data, the authors fit the asymptotic g-mode period spacing parameter $\Pi_0$ and near-core rotation $f_{\rm rot}$ against a grid of stellar models using a Mahalanobis-distance merit function within a Bayesian framework. They find that for some bright, well-behaved stars (e.g., KIC 4930889, 5941844, 7760680) TESS-based inferences can be compatible with Kepler results within $2\sigma$, though uncertainties are large and depend on data continuity and pattern completeness; for other targets, sparse patterns lead to biased or unconstrained parameters. The work highlights both the potential and the limitations of using TESS data for forward asteroseismology of gravity-mode pulsators, arguing for careful light-curve extraction, objective pattern identification, and acknowledgment of degeneracies in $\Pi_0$-driven modelling. Overall, the study demonstrates that TESS can complement Kepler in SPB asteroseismology but does not universally replace long-baseline, high-S/N data for precise interior-property inferences.

Abstract

The slowly pulsating B (SPB) stars are a class of variable star with masses between about 3 and 8 M$_{\odot}$. Their gravity-mode pulsation frequencies are sensitive to the near-core structure, which makes them useful probes of rotation and mixing in the deep stellar interior. Time series photometry, such as from the Kepler and TESS space telescopes, allows the extraction of their pulsation frequencies and construction of period spacing patterns. Previously, samples of slowly pulsating B stars were observed by the Kepler mission and underwent forward asteroseismic modelling to retrieve stellar parameters such as mass, age and core mass. However, all of these stars have since been re-observed by the ongoing TESS mission with light curves that are usually shorter and non-continuous, resulting in more difficult frequency extraction and interpretation in terms of constructing period spacing patterns. In this paper we compare the results of forward asteroseismic modelling of a sample of SPB stars using intermittent TESS light curve data to those based on long-duration Kepler light curves. We show how in some cases that the masses and core masses derived from only a few sectors of TESS data agree well with the 4-yr Kepler mission results, despite the stars having far fewer significant pulsation frequencies in their TESS light curves. However, some stars yield incompatible results, emphasising the complexities in forward asteroseismic modelling of gravity-mode pulsators with sparsely sampled or short duration TESS light curves.

Figures (15)

• Figure 1: Flowchart describing the light curve extraction process.
• Figure 2: Pixel rings used for iteratively determining each target's optimal aperture size, which are represented in different colours. In this schematic, the target star is located within pixel (10,10). Panel a shows the ring shape for a star located centrally within this pixel. If the star is closer to the pixel edge, the ring shapes are extended vertically, horizontally or both. Panels b, c, and d show ring shapes for stars located closest to the top edge, right edge and top right corner, respectively.
• Figure 3: TESS light curve extraction summary of sector 82 for KIC 7760680. The target aperture mask that maximises the S/N of the dominant pulsation mode is highlighted in red in the top two panels, with left- and right-hand panels showing flux and S/N, respectively. The panel below this shows the S/N curves of growth for the number of concentric rings in determining the optimal pixel mask (black) and the number of principal components (PCs, red dashed), with the optimal values circled. The bottom two rows show the PCs, with the first PC at the bottom of the panel, and the final detrended light curve for the sector.
• Figure 4: Percentage of frequencies in the Kepler period spacing patterns reported by pedersen2021 that were extracted in our analysis of TESS light curves. We were able to find period spacing patterns for the stars with filled symbols. The length and quality of the data was insufficient to find a convincing pattern in the TESS light curves for stars with open symbols.
• Figure 5: The evolution of the asymptotic period spacing, $\Pi_0$, with central hydrogen mass fraction, $X_{\rm c}$, for selected masses within the grid of structure models used in this work. The value of CBM, $f_{\rm CBM}$, is indicated by either solid or dashed lines. These models have $\log{D_{\rm ext}}=0$; higher $\log{D_{\rm ext}}$ models have similar $\Pi_0$ values.