Astron. Astrophys. 325, 1115-1124 (1997)

1. Introduction

Stars with spectral types later than approximately mid-F, and possibly as early as  A7, show signs of magnetic surface activity, accompanied by an outward temperature increase in their outer atmospheres. The amounts of radiation originating from different temperature intervals in these outer atmospheres (i.e., the chromosphere, the corona, and the transition region between these two) have been used to measure this magnetic activity; well-known examples are the chromospheric Ca II H&K line core emission and the coronal X-ray emission (for reviews see Zwaan 1991, and Vaiana 1983).

Several studies (e.g., Rutten et al. 1991, Zwaan 1991, and references therein) have shown that chromospheric and coronal emissions are strongly correlated for stars with spectral types from mid F to mid K. Main-sequence stars and evolved stars follow the same relationship between the stellar surface fluxes emitted in X-rays and in the Ca II H&K line cores, independent of spectral type, provided that a minimum flux density (which depends on spectral type, and perhaps weakly on luminosity class) is subtracted from the observed value.

Rutten et al. (1991) showed that the power-law index of a flux-flux relation increases with the difference between the formation temperatures of the two radiative diagnostics. Schrijver (1993) argues that the non-linearity of the stellar Ca II - X-ray relationship is caused by the non-linear dependence of Ca II K line-core emission on the mean magnetic flux density, as observed in solar active regions (Schrijver et al. 1989) and seen in model calculations by, e.g., Solanki et al. (1991), while the X-ray flux density is proportional to the magnetic flux density (Schrijver et al. 1987).

Studies on flux-flux relationships all result in the same quantitative description of these relationships, but they differ significantly on qualitative aspects as, e.g., which unit of radiative emission is best describing the flux-flux relationships; is there one flux-flux relation for all (`normal') magnetically active stars (from late A to M and from giants to dwarfs); what is the power-law index of the relationship (when the relationship can be expressed as a power-law)? The main reason why these studies do not give the same results is that the samples of stars used have not been the same and have been rather small (a few tens of stars). In order to verify the general validity of the relation between the variable coronal and chromospheric fluxes, independent of effective temperature and surface gravity, and in order to answer the questions stated above, it is necessary to derive this relation as accurately as possible for a large sample of stars. This is the first topic addressed in this paper.

Another question concerns the effect of variability on time scales larger than about one day on the scatter about the average relationship. Schrijver et al. (1992) found that the flux-flux relations between chromospheric and coronal diagnostics derived until then showed no deviations other than measurement uncertainties when the fluxes are observed with time intervals of only a few days or less. This was based on a small sample of 20 F6-K2 dwarfs and giants. With more accurate observations and a much larger sample of stars the contribution of variations in magnetic structure in the observed atmosphere to the remaining scatter can be studied. The deviation of an individual star from the mean X-ray vs. Ca II relationship is expected to depend on the time difference between the measurements.

The ROSAT All-Sky Survey, which was conducted from July 1990 until January 1991, offered an opportunity to determine the relationship between the X-ray and Ca II emission for a sizable sample of stars. We obtained Ca II H&K line photometry for 215 F-, G- and K-type stars at the Mt. Wilson Observatory at times as close as possible (mostly within a few days) to the ROSAT All-Sky Survey observation times. In Sect. 2 we discuss the observations and data reduction of the Ca II photometry, and of the X-ray measurements. Flux densities are derived in Sect. 3. The relationship between the stellar flux densities in X-rays and in the Ca II H&K lines is derived in Sect. 4. In Sect. 5 we discuss our results and compare them with previous work. Our conclusions are presented in Sect. 6.

© European Southern Observatory (ESO) 1997

Online publication: April 28, 1998

helpdesk.link@springer.de