LINE-PROFILE VARIATIONS IN M-GIANTS - CLUES TO MASS-LOSS AND CHROMOSPHERIC HEATING MECHANISMS

Citation
Pg. Judge et al., LINE-PROFILE VARIATIONS IN M-GIANTS - CLUES TO MASS-LOSS AND CHROMOSPHERIC HEATING MECHANISMS, The Astronomical journal, 105(5), 1993, pp. 1973-1986
Citations number
55
Categorie Soggetti
Astronomy & Astrophysics
Journal title
ISSN journal
00046256
Volume
105
Issue
5
Year of publication
1993
Pages
1973 - 1986
Database
ISI
SICI code
0004-6256(1993)105:5<1973:LVIM-C>2.0.ZU;2-U
Abstract
We analyze time-series, high dispersion spectra of the Mg II h and k a nd Ca II H and K lines of the semiregular giants p Per (M4 II-III, per iodicity P approximately 50 days), R Lyr (M5 III, P approximately 46 d ays), and g Her (M6 III, P approximately 90 days). The targets were st rategically selected to study the relative importance of convective mo tions and global stellar pulsations in determining the structure of th e outer atmospheres. These first or second ascent stars lie in a cruci al region of the HR diagram where circumstellar dust shells begin to f orm. We relate line profile variations to photospheric variations usin g the Fine Error Sensor on the IUE satellite and ground-based UBV phot ometry. We have detected small amplitude (typically +/- 10% ) but real changes in the profiles of Mg II and Ca II lines in all three stars w hich are not obviously related to underlying changes in the photospher es. We argue that the observed variability is due to changes in chromo spheric conditions and not variations within the circumstellar shell. The observations support the picture of a steady-state ''chromosphere' ' which is modulated on long time scales (approximately weeks). In g H er, we find evidence for a localized heating event which we tentativel y assign to the overshooting of a large supergranule cell. Unfortunate ly, some important conclusions of related work by Eaton et aL [ApJ, 36 4, 259 (1990)] warrant a re-examination in the light of our analysis. We conclude that the geometric scales over which most of the energy is deposited in the outer atmospheres must be much less than a stellar r adius, in contrast to large amplitude variables which are only slightl y more evolved.