INFERENCE OF STEADY STELLAR WIND V(R) LAWS FROM OPTICALLY THIN EMISSION-LINES - II - OCCULTATION EFFECTS AND THE DETERMINATION OF INTRINSICSTELLAR PROPERTIES

Citation
R. Ignace et al., INFERENCE OF STEADY STELLAR WIND V(R) LAWS FROM OPTICALLY THIN EMISSION-LINES - II - OCCULTATION EFFECTS AND THE DETERMINATION OF INTRINSICSTELLAR PROPERTIES, Astronomy and astrophysics, 330(1), 1998, pp. 253-264
Citations number
16
Categorie Soggetti
Astronomy & Astrophysics
Journal title
ISSN journal
00046361
Volume
330
Issue
1
Year of publication
1998
Pages
253 - 264
Database
ISI
SICI code
0004-6361(1998)330:1<253:IOSSWV>2.0.ZU;2-2
Abstract
This paper extends previous work on the inversion of line profiles to obtain wind velocity laws to a case that includes the occultation of l ight from the far side of the star. The velocity law v(r) is assumed t o be from a wind that is steady and spherically symmetric. The wind is also assumed to be optically thin in the emission line profile. The m ajor result here is the derivation of an analytic inversion formula. T he effects of stellar occultation are shown to produce a significant c hange in the analysis from paper I, and by accounting for the occultat ion, the red-shifted emission of P Cygni profiles can be used to obtai n v(r). Using simulated line profiles as generated from a radiation tr ansport code to test the procedure, the inversion technique based on o ptically thin lines successfully recovers v(r) distributions for weak LTE H-alpha profiles from hot star winds. Even in the case of NLTE H-a lpha lines, the technique is seen to reproduce the model velocity dist ribution quite well. Our inversion technique thus remains robust outsi de the scope of our assumptions, owing primarily to an empirical appro ach for applying the method. An important aspect of our empirical appr oach is the possibility of estimating intrinsic stellar and wind prope rties, such as the mass-loss rate M, photospheric radius R, and the st ellar distance D. As an example, photospheric stellar radii are derive d from the model profiles and found to be in good agreement with the i nput values, with typical errors of about 5%. Even in the NLTE case, t he photospheric radii are underestimated by only 10-20%.