A census of non-variable stars from Kepler and TESS data

www.postersession.com www.postersession.com www.postersession.com Figure 1.: The log g versus Teff diagram for Sample 1 (lower panel) and Sample 2 (upper panel). applied a simple five-sigma clipping to remove outliers. No further cleaning algorithms were applied to avoid removing any intrinsic variability. For our purposes, we used the Lomb-Scargle algorithm because it also includes a False-Alarm probability estimation. The method is a variation of the Discrete Fourier Transform, in which an unequally spaced time series is decomposed into a linear combination of sinusoidal and cosinusoidal functions. Based on the instrumental characteristics and the known periods of variable stars, we selected three different frequency domains for which we calculated the Lomb-Scargle periodograms and the False Alarm probabilities for the highest peak. These are: •Domain 1: below 0.1 c/d •Domain 2: 0.1 to 2.0 c/d •Domain 3: 2.0 to 25.0 c/d Finally, we have defined two samples: •Sample 1: due to the discussed possible instrumental effects in the low-frequency domain, this sample consists of 21 734 stars with log FAP ≥ −2 in the other two domains (0.1 to 25.0 c/d). •Sample 2: this stricter set comprises 3265 stars for which log FAP ≥ −2 is true for all three frequency domains We investigated the locations of the stars in both final samples in the HRD, as shown in Figure 1. All evolutionary stages up to the red giant branch (RGB) are well populated. 1. Paunzen et al., 2024, A&A, 687, A208 One of the essential questions related to studies aimed at describing stellar objects is whether all of them are variable. The most commonly given answer is that all stars are variable, but it is only the amplitude that makes a significant difference. The Sun, the closest star to our planet, is a perfect example. Its variability amplitude depends on the wavelength region (with more considerable changes at shorter wavelengths). Furthermore, a timescale of approximately 11 years, with an amplitude of about one millimag (mmag), a 27-day rotation period (resulting in a change of two mmag), and five-minute short-scale variations (of 0.15 mmag) has been identified. An observer from the outside would measure the superposition of all these variations integrated over the solar surface. It also demonstrates the importance of time sampling in detecting variations on different time scales. In general, variable stars are at the centre of many scientific studies. For example, Cepheid variables and their period-luminosity relations have enabled us to begin constructing a distance ladder, which has helped us explore large regions of the Universe. The periods, amplitudes, and light curve characteristics are as manifold as the underlying physical mechanisms. However, it is essential not to lose sight of non-variable stellar objects, as they are very much needed for calibrating absolute fluxes and radial velocities (most photometric variable stars also show substantial spectroscopic variations). Techniques such as fitting the spectral energy distribution are crucial, depending on the non-variability. Many photometric calibrations of the effective temperature and metallicity are based on Galactic field stars. However, the variability of the individual stars is generally neglected. We first selected all stars with long cadence light curves available within the original Kepler mission for our analysis. We did not consider the about 2000 short cadence light curves because they are only a small fraction of the overall data set. Furthermore, we have selected isolated stars of the Golden Gaia OBAF Sample and extracted their TESS data. As the next step, we removed all stars already known as variables. For this, we utilised the catalogue of the Asteroid Terrestrial-impact Last Alert System (ATLAS) and the International Variable Star Index (VSX), along with several works that present automatic variable star detection routines using Kepler and TESS data. No further constraints were set. The final sample, further processed, consisted of about 250 000 light curves. As a next step, we investigated the location of the target stars in the Hertzsprung-Russell diagram (HRD). For our analysis, we began with the PDCSAP fluxes from the Kepler mission, using only data with optimal quality flags. The different quarters of one data set have different offsets, which we initially corrected. For the TESS data, we only took isolated stars. We then A census of non-variable stars from Kepler and TESS data E. Paunzen1, L. Kueß2, K. Neumannová1, P. Mondal1 1Department of Theoretical Physics and Astrophysics, Masaryk University, Brno, Czechia (contact: [email protected]) 2Department of Astrophysics, Vienna University, Austria Bibliography Introduction Abstract: The analysis of non-variable stars is mostly neglected in the literature. However, such objects are needed for many calibration purposes and for testing pulsational models. The photometric time series from the Kepler satellite mission are still the most accurate available and are excellently suited for searching for non-variable stars. Furthermore, the TESS satellite mission brought a new variety of light curves for millions of objects. We analysed all Long Cadence light curves for stars not reported as variables so far from the Kepler satellite mission, and all stars in the classical instability strip from the TESS mission. Using the known characteristics and flaws of these data sets, we defined three different frequency ranges in which we searched for nonvariability. We used the Lomb-Scargle periodogram and the False-Alarm probability (FAP) to analyse the cleaned data sets in the ranges of below 0.1 c/d, 0.1 to 2.0 c/d, and 2.0 to 25.0 c/d, respectively. Furthermore, we also calculated the standard deviation of the mean light curve to provide another parameter. In this poster, we present the results of our analysis. Target selection Results Data preparation