Nikki Miller
Career Stage
Student (postgraduate)
Poster Abstract
Stars with accurate and precise effective temperature measurements are needed to test stellar atmosphere models and calibrate empirical temperature estimates, but there are few such suitable standard stars. Parallaxes from Gaia now make it possible to measure the fundamental effective temperature of many stars in detached (non-interacting) eclipsing binaries, and establish a set of solar-type benchmark stars with accurate and precise temperatures. We developed a new method for calculating the temperature of eclipsing binary stars that uses high precision photometry, parallax and multi-wavelength photometry. We applied it to a well studied system, AI Phoenicis (AI Phe), and obtained robust temperature measurements of 6199 ± 22 K for the F7V component and 5094 ± 16 K for the K0IV component.
Plain text summary
Abstract. Stars with accurate and precise effective temperature measurements are needed to test stellar atmosphere models and calibrate empirical temperature estimates, but there are few such suitable standard stars. Parallaxes from Gaia now make it possible to measure the fundamental effective temperature of many stars in detached (non-interacting) eclipsing binaries, and establish a set of solar-type benchmark stars with accurate and precise temperatures. We developed a new method for calculating the temperature of eclipsing binary stars that uses high precision photometry, parallax and multi-wavelength photometry. We applied it to a well studied system, AI Phoenicis (AI Phe), and obtained robust temperature measurements of 6199±22K for the F7V component and 5094±16K for the K0IV component.
Method. To measure the fundamental effective temperature, we need to know the radius of the stars, emitting a total flux in all wavelengths, measured at a certain distance from the observer.
The radius can be obtained by fitting light curves with eclipsing binary models. With TESS observations, the radii of some stars can be measured with uncertainties smaller than 0.2%! Figure shows the primary and secondary eclipses of AI Phe taken by TESS, with a very good light curve model.
How do we get the total flux? We aim to balance contributions from data and physics by creating a flux “integrating function” with realistic small-scale features like absorption lines but a broad shape determined by observations. This is done by using polynomials to distort model spectral energy distributions (two blackbody curves shown) for each star to “fit” multi-wavelength magnitude, colour and flux ratio observations, along with other data and constraints. Also plotted are response profiles of different filters, and flux ratio data increasing from 0 in ultraviolet to 2 in infrared.
The distance is obtained from Gaia parallaxes.
Results. We applied our method to the well-studied eclipsing binary AI Phe. This is a F7V + K0IV binary (ideal benchmark stars) with good quality light curves in several photometric bands from near IR to near UV. It has very accurate radii from Maxted et al (2020) and a strong upper limit on interstellar reddening. Plotted in log surface gravity-log temperature space are AI Phe, the Kepler LEGACY sample of stars with detailed asteroseismology, and 8 Gaia FGK-type benchmark stars. AI Phe lies some distance from other benchmark stars. Another figure shows output flux “integrating functions” which are modified versions of the spectral energy distributions shown before. Below are the distortion functions, showing how much the integrating functions were distorted as a function of wavelength. Most distortion occurs where there is not much flux, but is significant in UV. We measured T1=6199±22K; T2=5094±16K. We tested all aspects of the method until we were satisfied our results were robust.
Next steps and prospects. More stars! We will continue applying the method to suitable eclipsing binary stars. There are already a handful of targets from K2 fields but we expect many more to be observed by TESS over the next few years. Follow up observations! We recently won observing time on the UVES instrument on the VLT to follow up our work on AI Phe. By taking a spectrum during a total eclipse we can separate the spectra of the two stars for an independent determination of temperature and chemical composition. Benchmarks! These results contribute to the PLATO Benchmark Stars work package, which aims to “measure precise and accurate fundamental properties of stars that can be used to validate and improve data products from the PLATO mission”. PLATO is an upcoming ESA mission which will search for Earth-like planets in the habitable zone around solar-type stars.
Method. To measure the fundamental effective temperature, we need to know the radius of the stars, emitting a total flux in all wavelengths, measured at a certain distance from the observer.
The radius can be obtained by fitting light curves with eclipsing binary models. With TESS observations, the radii of some stars can be measured with uncertainties smaller than 0.2%! Figure shows the primary and secondary eclipses of AI Phe taken by TESS, with a very good light curve model.
How do we get the total flux? We aim to balance contributions from data and physics by creating a flux “integrating function” with realistic small-scale features like absorption lines but a broad shape determined by observations. This is done by using polynomials to distort model spectral energy distributions (two blackbody curves shown) for each star to “fit” multi-wavelength magnitude, colour and flux ratio observations, along with other data and constraints. Also plotted are response profiles of different filters, and flux ratio data increasing from 0 in ultraviolet to 2 in infrared.
The distance is obtained from Gaia parallaxes.
Results. We applied our method to the well-studied eclipsing binary AI Phe. This is a F7V + K0IV binary (ideal benchmark stars) with good quality light curves in several photometric bands from near IR to near UV. It has very accurate radii from Maxted et al (2020) and a strong upper limit on interstellar reddening. Plotted in log surface gravity-log temperature space are AI Phe, the Kepler LEGACY sample of stars with detailed asteroseismology, and 8 Gaia FGK-type benchmark stars. AI Phe lies some distance from other benchmark stars. Another figure shows output flux “integrating functions” which are modified versions of the spectral energy distributions shown before. Below are the distortion functions, showing how much the integrating functions were distorted as a function of wavelength. Most distortion occurs where there is not much flux, but is significant in UV. We measured T1=6199±22K; T2=5094±16K. We tested all aspects of the method until we were satisfied our results were robust.
Next steps and prospects. More stars! We will continue applying the method to suitable eclipsing binary stars. There are already a handful of targets from K2 fields but we expect many more to be observed by TESS over the next few years. Follow up observations! We recently won observing time on the UVES instrument on the VLT to follow up our work on AI Phe. By taking a spectrum during a total eclipse we can separate the spectra of the two stars for an independent determination of temperature and chemical composition. Benchmarks! These results contribute to the PLATO Benchmark Stars work package, which aims to “measure precise and accurate fundamental properties of stars that can be used to validate and improve data products from the PLATO mission”. PLATO is an upcoming ESA mission which will search for Earth-like planets in the habitable zone around solar-type stars.
Poster file
Poster Title
Accurate and precise effective temperature measurements of eclipsing binary stars
Url
nikkimiller.space