J.L. Linsky, B.E. Wood, A. Brown, and R.A. Osten
We report on moderate (lambda/dlambda = 20,000) and high (lambda/dlambda = 90,000) resolution spectra of the 104 day period Capella binary system (HD 34029) obtained with HST's Goddard High Resolution Spectrometer (GHRS) on 1995 September 9 and 1996 April 9. The observations include a long-duration moderate resolution spectrum of the coronal FeXXI 1354 A line and both moderate and high resolution spectra of the HeII 1640 A multiplet.
Our objective in observing the FeXXI line formed at T = 1E+07 K is to determine for the first time the line shape parameters and contribution of each star's corona. This is feasible because the GHRS can resolve the 53 km/sec radial velocity separation of the stars. Our analysis led to four surprising results: (1) The contribution of the slowly-rotating G8III star to the total FeXXI 1354 A flux is similar to that of the more active rapidly-rotating G1III star, in contrast to other UV lines formed at lower temperatures. (2) The centroid velocities of the FeXXI lines from both stars are near their respective photospheric radial velocities. Thus, there is no evidence for downflows or winds, and the hot coronal plasma must be confined, presumably by strong, closed magnetic fields. This is the first direct kinematic evidence for magnetic confinement in the corona of a giant star. (3) The line widths are thermal, indicating very low turbulence (xi < 23 km/sec) compared to the 54 km/sec thermal speed. (4) Our analysis of EUVE spectra that include four FeXXI lines shows that the volume emission measures of these lines are about a factor of 3 smaller than for the 1354 A line, which was observed several months after the closest EUVE observation. We consider possible explanations for this discrepancy and conclude that variability of the high temperature coronal emission is the most likely explanation.
We observed the HeII 1640 A line twice with a time separation of 7 months, corresponding to about 2 orbital periods. The profiles are nearly identical, except for a large difference in flux near the expected radial velocity of the G8III star. We believe that the broad HeII emission produced by the G1 star is formed mostly by collisional excitation in its transition region, while the emission from the G8 star is produced predominantly by a photoionization/recombination process driven by EUV radiation from the G8 star's corona. The HeII 1640 A flux that we measure from the G8 star is consistent with predictions based on EUVE flux measurements. The decrease in the 1640 A emission from the G8 star between the phase 0.73 and 0.78 measurements indicates that its EUV radiation is variable as is observed by EUVE.