Time-frequency analysis of δ Scuti light curves using the wavelet transform.
δ Scuti stars are intermediate-mass stars (i.e. between 1.5 and 3 solar masses) pulsators located in the classical Cepheids instability strip, either in their pre- main sequence stage, or already in the main sequence or also moving from the main sequence to the giant branch. Their spectral types range from A to F. Before the era of space telescopes, these stars were thought to have a small number of pulsating frequencies, but with the data coming from space missions like CoRoT, Kepler, TESS, etc., it has been found that there are some δ Scuti stars with hundreds of pulsation frequencies. In the last years, some authors found an amplitude modulation in a few δ Scuti stars but a time-frequency analysis has not yet been performed up to date. In our work we have tried the continuous wavelet transform as well as a multi resolution analysis using wavelets as filters. Unlike the short time Fourier transform, that uses a fixed windows size, the wavelet transform uses short windows at high frequencies and long windows at low frequencies, making it like an adaptive analysis tool. The wavelet analysis relies on the use of a mother wavelet, i.e. a wave-like function of finite energy that can be scaled and shifted in a way that conserves the energy, so it can correlate with the variations o the signal, thus providing a measure of the instantaneous frequencies of the signal. The continuous wavelet transform is widely used in sound processing and pattern recognition applications, making it a suitable tool to use with light curves. We show this analysis applied to a sample of δ Sct stars and we try to give an explanation to the pulsating frequencies that show irregularities.
Age constraints for the open cluster NGC 2477 using asteroseismology from TESS FFI
Precise ages of stars are important for understanding stellar
and galactic evolution. Asteroseismology can provide precise ages of
stars. Our objective is to combine NASA TESS data along with the ESA
GAIA photometric data, to search for, to detect and to analyse
oscillations in cluster members. Since stars in a cluster are assumed to
have the same age and chemical composition, this provides constraints
for the interpretation of asteroseismic data. We analysed the
intermediate-age open cluster NGC 2477, known to suffer from
differential extinction. Using TESS FFI images, we identified several
levels of variability of stars in this cluster, including binaries and
short- and long-period oscillating stars. We perform an asteroseismic
analysis on a small set of uncontaminated oscillating stars. We used
MESA and GYRE to make stellar models and to calculate the oscillation
frequencies. In this talk, I will present how we constrained the
cluster’s age by asteroseismic modelling of oscillating stars using the
general properties of the cluster along with the frequencies from
oscillating stars.
The study of stellar granulation has vastly improved thanks to the observations made by Kepler, which have shown that the stellar flux produces stochastic brightness fluctuations. These fluctuations need to be quantified in order to detect and characterize exoplanets. In this work, we aim to provide new scaling relations between granulation properties, such as the rms brightness fluctuation and characteristic timescale, with stellar parameters such as the peak frequency, using long time series of 3D stellar atmosphere models at different metallicities and across the HR diagram, generated with the STAGGER-code. To validate our results, we compared our theoretical granulation properties with the values of a large sample of Kepler stars from the literature, and analyzed selected stars with accurate stellar parameters as well. In this talk I will present the relations that our 3D models have with stellar parameters and the agreement with observational data.
Artificially increasing the luminosity and the thermal diffusivity of a model is a common tactic adopted in hydrodynamical simulations of stellar convection. In this work, we analyse the impact of these artificial modifications on the physical properties of stellar interiors and specifically on internal gravity waves (IGW). We perform two-dimensional simulations of solar-like stars with the fully compressible MUltidimensional Stellar Implicit Code (MUSIC). We compare three models with different luminosity enhancement factors to a reference model. The results confirm that properties of the waves are impacted by the artificial enhancement of the luminosity and thermal diffusivity. We find that an increase of the stellar luminosity yields a decrease of the bulk convective turnover timescale and an increase of the characteristic frequency of excitation of the internal waves. We also show that a higher energy input in a model, corresponding to a larger luminosity, results in higher energy in high frequency waves. Across our tests with the luminosity and thermal diffusivity enhanced together by up to a factor of 10^4, our results are consistent with theoretical predictions of radiative damping. Increasing the luminosity also has an impact on the amplitude of oscillatory motions across the convective boundary. One must thus interpret with caution studies of IGW based on hydrodynamical simulations with artificially enhanced luminosity.
Although seismic inference has proven to be quite valuable when determining stellar parameters of solar-like pulsators, estimating these very parameters can be subject to biases and thus remain uncertain. A particularly important example is the helium-mass degeneracy, where the uncertainties regarding the internal physics cause a poor determination of both the mass and surface helium content. Accordingly, an independent helium estimate is needed to overcome this degeneracy. A promising way to obtain such an estimate is to exploit the so-called ionisation glitch, a deviation from the asymptotic oscillation frequency pattern caused by the rapid structural variation within the ionisation zone. However, although becoming more and more sophisticated, current approaches for modelling glitches face problems such as the need for calibration by realistic stellar models. Accordingly, we focused on an alternative approach to studying glitches using a physically motivated model of the ionisation region. This is done by means of an analytical approximation of the first adiabatic exponent inside an adiabatically stratified region and the derived structure is found to depend on three parameters. They respectively control the surface helium abundance, the electron degeneracy in the convective zone, and the extent of the ionisation region. Characterising the glitch based on such a model would thus allow us to extract physical quantities such as the helium abundance without the need for calibration, as well as other interesting constraints such as the electronic degeneracy in the star’s convection zone.
The main goal of my studies is to demonstrate how the binary evolution affects the present stage and pulsational properties of δ Scuti stars. To do so, I perform an in-depth analysis of all available photometric data and spectroscopic observations for studied binaries, KIC 10661783 and AB Cas. In order to find models describing the past evolution and current observational status of the analysed systems, I construct binary evolution MESA models that reproduce the orbital periods, masses and radii of binary systems. During my seminar I will focus on the properties of calculated evolutionary models which may have a significant impact in the future prospects of the δ Scuti analysis.
Since the first electronic computers became available, astronomers have applied them to model the structure and dynamics of stars. The first attempts neglected convection and turbulence, considering only radiative energy transport. However, it turned out soon that we could not adequately describe pulsation without convection. Moreover, the different improved forms of static mixing length theory were also inadequate. Hence massive research was started to create a time-dependent theory that can describe convection correctly in a one-dimensional approximation. These efforts unfolded some hidden features of phenomena but could not answer all of the questions raised. Since convection and non-radial pulsation are genuinely multi-dimensional phenomena, multi-D models seem inevitable, but this approach requires high computational performance, which was not available decades ago. In addition, the multi-dimensional treatment of the problems allows us to directly study the interaction between convection and pulsation. Besides this advantage, we can also test and calibrate our 1D approaches and understand the connected physical transport processes more deeply and precisely. In the seminar, we will compare two extensively used 1D pulsation models and a multi-dimensional one in detail in case of classical pulsating variable stars. We show two approaches to calibrate our 1D models. Firstly, we calibrate the models to the observational constraints to have our best possible parameter set. Secondly, we can test these 1D systems by the multi-dimensional models, finding out whether the usual conventions of time dependent mixing length theory are valid or not. These calibrations and tests will help us to understand better the multidimensional nature of complex physical phenomena, and also to fully exploit the large amount of photometric and spectroscopic data coming from ongoing and planned large sky surveys.
Studies of the Sun’s interior through the analysis and interpretation of its p-mode oscillations — helioseismology — have illuminated a vast and diverse range of physical and astronomical phenomena. Aside from the Sun, measurements of solar-like p-modes are easiest for post-main-sequence stars; subgiants and red giants now constitute the majority of our seismic observations. However, foundational assumptions underlying helioseismology cease to be applicable for these evolved stars, on account of their different internal structures, rendering them impenetrable to analysis of a similar kind. Whereas the helioseismic techniques in question borrow heavily from the quantum mechanics of atomic systems, I will demonstrate that oscillations in these giants — which possess mixed p-mode and g-mode character — behave like acoustic “molecules”, rather than atoms, and therefore demand the adaptation of techniques from quantum chemistry instead. I describe two applications of this construction to constraining stellar properties by forward modelling. First, I will show that it permits the correction of systematic errors resulting from deficiencies in numerical models of stellar surfaces. I will also use it to describe the nonlinear effects of radial differential rotation on these mixed modes: left unaccounted for, these may yield erroneous diagnoses and measurements of buried magnetic fields.
During the Milky Way’s lifetime, it has merged with several dwarf galaxies. One of the largest mergers was with the galaxy know as Gaia-Enceladus-Sausage (GES). To fully understand the evolution and current structure of our Galaxy it is important to know when these mergers happened. Using asteroseismology in combination with spectroscopic and photometric measurements of solar-like oscillators we can determine the ages of the stars with great accuracy. In this talk, I present how we use Gaia eDR3 data to make a dynamical selection of red giant stars that once belonged to the GES galaxy. I also describe how we use asteroseismology from the Kepler and K2 mission along with chemical abundances from APOGEE and photometry from 2MASS to determine the ages of these GES stars. From these age determinations, we can constrain when the GES merger happened and thereby contribute to mapping out the evolutionary history of our Galaxy.
Mechanisms governing angular momentum transport remain one of the main problems to a better understanding of the evolution of main-sequence solar-like stars. Their rotational evolution along the main sequence will remain an open question until we are able to provide reliable observables of both the surface and the core rotation rate for such targets. While p-mode rotational splittings and photometric follow-up of the surface brightness allow independent measurements of stellar rotation rate on the upper regions of these stars and their surface, only g modes are able to provide information about the rotational state of deeper regions. Reliable characterisation of g modes in main-sequence solar-like star would thus be a great step forward concerning our understanding of stellar-interior dynamics. In this talk, I will first describe ROOSTER, a machine learning methodology with ~97% accuracy, designed to reduce the amount of visual inspections when extracting rotation periods from Kepler light curves modulations. I will especially comment on the possibility to compare those results with rotation periods yielded by the analysis of p-mode rotational splittings. Moreover, I will present new perspectives to observe low-degree, low-order solar p modes using the Solar-SONG échelle spectrograph. I will explain how the reduction and analysis I performed on the data collected during the 2018 observational run allows us to improve the SNR below 2 mHz compared to a space instrument like GOLF or a ground-based network like BiSON. I will finally present results from 3D deep-shell hydrodynamical simulations of a F-type star with the ASH code, in which a thin external convective envelope is coupled with the radiative interior. I will show that the differential rotation state of F-type star is different from what is observed in cooler stars. I will then confront the g-mode excitation and visibility at the surface in the simulation with former results obtained for the Sun. Because of the shape of the Brunt-Väisälä profile, the range of excited g modes also significantly differs from what is observed in solar models.