Research projects

Geometry and evolution of stellar magnetic fields

The origin and role of magnetic fields in the evolution of stars is the last major unsolved problem of stellar physics. Astronomers agree that magnetism plays a central role at many stages of stellar life: from the formation of stars and planetary systems to supernovae explosions and winds of dying stars. Magnetic activity of the Sun is clearly observed as 11-year sunspot cycle that affects the entire solar system.

Despite an abundance of indirect evidence of the importance of magnetic fields, direct detection and study of magnetic fields on stars other than the Sun is an extremely challenging task. Such analysis was impossible until recently, leading astronomers to speculate about stellar magnetism based on theoretical models extrapolated from the Sun using unreliable indirect proxy indicators (flaring, X-ray emission).

During the last decade new instruments at large telescopes, especially high-resolution spectropolarimeters, have allowed detecting magnetic field in essentially all types of stars. These discoveries led to a completely new and, often, unexpected perspective on the geometry and evolution of stellar magnetic topologies. Theoretical studies are only now starting to explore implications of this new picture of stellar magnetism.

The main topic of my research is observational investigation of magnetic fields and reconstruction of magnetic field topologies in massive, intermediate-mass and low-mass stars. In particular, I carry out the following research projects:

Spectropolarimetric studies and modelling of magnetic field topologies in massive O and B stars
Four Stokes parameter observations and reconstruction of magnetic field topologies in magnetic Ap/Bp stars
Polarimetric detection and mapping the structure of magnetic fields in young active Sun-like stars
Study of magnetic fields in low-mass stars (M dwarfs and T Tauri stars) using high-resolution infra-red spectroscopy

a2CVn MDI Magnetic field structure of the Ap star α2 CVn reconstructed using spectropolarimetric observations in all four Stokes parameters.
Top panel: field strength distribution
Bottom panel: orientation of magnetic field vectors

Stellar non-radial pulsations

Many stars change periodically their brightness and radius due to pulsational instability. Oscillations provide unique possibility to measure properties of the stellar interiors and determine fundamental parameters of stars using a technique known as asteroseismology.

Magnetic Ap stars represent are among the most interesting targets for applications of asteroseismology. These stars exhibit non-radial p-mode oscillations with periods 5-15 minutes in the presence of highly non-uniform atmospheric chemistry and strong magnetic fields. I have pioneered spectroscopic time-resolved observations of pulsating Ap stars using large telescopes. These studies revealed a puzzling pulsational behaviour, characterized by a large diversity of radial velocity amplitudes and phase, and an unusual profile variations of the lines of rare-earth elements.

We have succeeded in interpreting and modeling pulsations in magnetic Ap stars in terms of magneto-acoustic waves, propagating outwards in stellar atmospheres with increasing amplitude and changing phase. This pulsational behaviour allows reconstructing from observations a three-dimensional picture of pulsation waves, chemical stratification, and magnetic fields, as can be done for no other star but the Sun.

roAp LPV Typical rapid line profile variations in an oscillating Ap star.
Left panel: the average spectrum and time evolution of the residuals showing pulsation in rare-earth lines.
Right panel: pulsational radial velocity curves showing the outward propagation of the magneto-acoustic wave.

Doppler Imaging of stellar surfaces

Doppler imaging (DI) is a powerful technique which allows deriving a two-dimensional map of the stellar surface from time-series observations of the line profile variations of rotating spotted star. On the one hand, DI uses detailed physical models of the stellar surface to calculate theoretical spectra of magnetic spotted stars. On the other hand, this method relies on regularized inverse problem solution algorithms. Currently, DI is the most power astronomical remote sensing technique. The (indirect) resolution of the stellar surfaces provided by this method by far surpasses possibilities of the largest existing or planned interferometers.

I am part of the world's leading group in imaging of stellar surfaces. I have developed Doppler and Zeeman-Doppler inversion codes for mapping of
chemical and isotopic spot distributions in Ap/Bp stars
temperature spot distributions in active late-type stars
magnetic field structure in both early-type and late-type stars
velocity field in non-radially pulsating stars

Using this software, I have carried out, for the first time, reconstruction of the stellar magnetic field geometries from the spectropolarimetric observations in all four Stokes parameters.

A brief overview of the scientific results obtained in our DI studies and an illustration of the basic principles of the DI technique can be found in my DI Gallery.

HR 3831 Chemical spot distribution at the surface of Ap star HR 3831.

Atmospheres of early-type stars

Atmosphere is a thin outer layer of the star where stellar radiation is formed. Models describing the temperature, pressure, and velocity field in stellar atmospheres are basic tools of the astronomical research on individual stars, stellar clusters and galaxies.

Standard 1-D theoretical codes are available to model atmospheres of cool and hot stars, assuming solar or scaled-solar chemical composition. Such tools are generally not applicable to B, A, and F stars with anomalous surface abundances. In these stars line opacity differs considerably from normal stars and also varies substantially from one star to another, depending on individual chemical composition and occasional presence of strong magnetic fields.

Together with my collaborators, I utilize an advanced opacity sampling stellar model atmosphere code LLMODELS to study effects of non-solar abundances, modification of the line opacity due to Zeeman splitting, magneto-hydrostatic equilibrium with the Lorentz force included, and an inhomogeneous vertical and horizontal distribution of chemical elements.

My theoretical work with the LLMODELS code is coupled with observational investigations using world's largest telescopes. These studies are aimed at testing new model atmospheres, finding and interpreting the signatures of anomalous atmospheric structure in chemically peculiar B-F stars.

aCir SED Spectral energy distribution of the cool Ap star alpha Cir modeled taking into account non-solar abundances and vertical stratification of chemical elements.

Chemical stratification in stellar atmospheres

Stellar chemical abundance analysis is the standard tool for measuring concentrations of various elements in the atmospheres of stars of different types. Information about stellar chemistry allows to probe Galactic chemical evolution and to test different scenarios of element production by thermonuclear reactions in stellar interiors and by supernovae.

The standard assumption of the abundance analyses is that stellar atmosphere is chemically homogeneous. Although reasonable for cool stars with convective envelopes, this may not apply to hotter stars without surface convection zones. The latter stars can possess considerable abundance gradients within the line forming regions due to operation of chemical transport processes.

Taking advantage of the wealth of information available in modern high-quality stellar spectroscopic observations, I have pioneered development and application of the inversion techniques to determine radial distribution of chemical elements in stellar atmospheres. My studies have added a new dimension to stellar abundance analysis, revealing, for the first time, vertical separation of elements in the atmospheres of chemically peculiar B-F stars.

HR 1217 Radial distribution of chemical elements in the atmosphere of late-A star HR 1217. Chemical abundances are given on the logarithmic scale relative to the Sun.

Spectroscopic studies of binary stars

Binary stars offer unique possibilities for testing the theories of stellar structure and evolution. In particular, the masses and radii of eclipsing spectroscopic binaries can be determined completely independently of theoretical assumptions. Binaries with components of similar mass are also very useful for studying the processes which change the chemistry of the stellar surface layers.

I have developed codes for simulating composite spectra of multiple stars and for disentangling contributions of individual stars using time-series spectroscopic observations. These techniques are applied in the context of chemical abundance studies of early-type binary stars. A puzzling result of these studies is existence of significant difference in the chemical composition of close binary components with very similar fundamental parameters.

AR Aur Disentangling of composite spectra of the eclipsing binary star AR Aur. The primary component of this system is a spotted chemically peculiar star of the rare HgMn-type.

Spectropolarimetric and spectroscopic observations of stars

I use spectrometers at the world's most sophisticated 8-m Very Large Telescopes to obtain spectroscopic observations of stars. These observational projects include high time resolution studies of the stellar non-radial pulsations using UVES, IR spectroscopy of low-mass stars with CRIRES, and polarization measurements with FORS 1/2.

I also employ high-resolution spectropolarimeters at the 2-4-m telescopes, in particular CFHT and NOT, to carry out investigations of stellar magnetic fields and to study inhomogeneities and dynamical phenomena at stellar surfaces.

A consortium including our group at Uppsala university has recently developed full Stokes vector polarimeter, HARPSpol, for the HARPS instrument at the ESO 3.6-m telescope. I am leading science verification observations and exploitation of the truly unique polarization data delivered by this most advanced stellar polarimeter.

VLT HARPS hr1099

Left: ESO 8-m Very Large Telescopes are our primary tools for obtaining precise observations of stellar spectra in the optical and IR wavelength regions.
Center: polarization spectrum recorded with the new four Stokes parameter polarimeter HARPSpol at the ESO 3.6-m telescope.
Right: circular and linear polarization detected for the RS CVn star HR 1099 using HARPSpol.

Computer programs for stellar physics research

I develop and maintain a variety of computer programs for stellar physics research. This includes codes for the visualization and analysis of spectroscopic observations, programs for calculation of theoretical stellar intensity and polarization spectra, and inversion codes for reconstruction of chemical stratification and horizontal maps of star spots and magnetic fields. Find out more details here.

I use existing and develop my own computer programs for 3-D visualization of the models of magnetic fields, non-radial pulsations, and surface inhomogeneities at the surfaces of stars.

giant hd37776 Left: density distribution in 3-D hydrodynamic simulation of convection in red supergiant star.

Right: 3-D rendering of the magnetic field structure in the He-strong Bp star HD 37776.

My research projects are supported by KAW KVA VR UU