Surface brightness distributions of late-type stars

Abstract

The aim of this work has been to increase our understanding of the surface brightness distributions of late-type stars through Doppler imaging and eclipse mapping techniques. Combining spectroscopic and photometric observations with the technique of Doppler Imaging, I have reconstructed surface images of the G2V star He 699 (for 08 October 2000), which show high latitude and polar structures. In the case of the KOV star AB Dor, the Doppler images for January 1992 and November 1993 show a large polar cap with small dark features also present at intermediate to high latitudes. As the phase sampling of the observations was insufficient to apply the sheared-image method it was not possible to detect any differential rotation.

In the second part of my thesis I determine the surface brightness distribution of the primary component of the RS CVn eclipsing binary SV Cam. I have used extrapolated size distributions of sunspots to an active star to synthesize images of stellar photospheres with high spot filling factors. The resulting surface images, reconstructed with the Maximum Entropy eclipse mapping technique, show large spurious spot features at the quadrature points. It is concluded that two-spot modelling or chi-squared minimisation techniques are more susceptible to spurious structures being generated by systematic errors, arising from incorrect assumptions about photospheric surface brightness, than simple Fourier analysis of the light-curves.

Spectrophotometric data from 9 HST orbits, observed in November 2001, have been used to eclipse-map the primary component of SV Cam. In combination with its HIPPAR- COS parallax it is found that the surface flux in the eclipsed low-latitude region is about 30% lower than computed from the best fitting PHOENIX model atmosphere. This flux deficit can only be accounted for if about a third of the primary's surface is covered with unresolved spots. However, when the spottedness from the eclipsed region is applied to the entire surface of the primary star, there still remains an unaccounted flux deficit. The remaining flux deficit is explained by the presence of a large polar spot extending down to latitude 48+/-6 deg. When the Maximum Entropy eclipse mapping technique is used to fit SV Cam's lightcurve, the observed minus computed residuals show strong spurious peaks at the quadrature points. It is only possible to reduce these peaks with the addition of a polar cap and the reduction of the primary star's temperature, to account for the star being peppered with unresolvable spots. Motivated by this result we investigate the limb darkening of the primary component of SV Cam. The wavelength dependence of the limb darkening is analysed by sub-dividing the HST lightcurve into 10 bands of equal emission flux. Flux variations between the first and fourth contact of the primary eclipse indicate that the limb darkening decreases towards longer wavelengths, in accordance with published limb darkening laws. Comparing fits of ATLAS and PHOENIX model atmospheres we find a wavelength dependence of the best fitting model. Due to its smooth cutoff at the stellar limb, the spherical geometry of the PHOENIX model atmosphere gives the best fit during partial eclipse. Between the second and third contact the difference between spherical and plane-parallel geometry is less important.