Astronomy and Astrophysics, volume 292, 191-207 (1994/12-1)
Applicability of the Rossby number in activity-rotation relations for dwarfs and giants.
Abstract (from CDS):
Empirical arguments for using the Rossby number as a rotation measure of active stars are reviewed using the log activity - rotation period diagrams. It is shown that when giant stars are considered the scatter on such diagrams cannot be appreciably diminished by replacing rotation period with the Rossby number. The observed scatter is, however, substantially reduced in activity-rotation diagrams of the solar type stars, i.e. dwarfs with 0.5≲B-V≲0.8, when the rotation period is scaled with appropriate factors, in other words, when the Rossby number is used. In case of active dwarfs lying outside of this spectral range, the using of the Rossby number does not improve the scatter. The scaling factors (called empirical turnover times) can be determined for different spectral types by minimizing the observed scatter. It is shown that the empirical turnover times determined from the calcium emission flux, magnesium emission flux and X-ray flux are identical with one another apart from constant factors: after a rapid increase with increasing F and early G spectral type, the turnover time levels off for late G and K types. The X-ray data indicate its second possible rise for M stars. Other available activity indices are useless for determination of the turnover time for M stars. Numerical relations between the calcium emission flux, X-ray flux or the surface magnetic field of active dwarfs on one side with the stellar color index and the Rossby number on the other side are determined from the observational data. Empirical formulae connecting the chromospheric and coronal radiative losses with the effective temperature and the average strength of the surface magnetic field (which is, by definition, proportional to the filling factor, hence depends on rotation) can be derived from the above relations. They indicate that the chromospheric losses are proportional to (Teff)2.9(<Bsurf>)0.6 whereas the coronal losses are more sensitive to the both parameters and are proportional to (Teff)8.3(<Bsurf>)1.9.