Ultraviolet interstellar linear polarization. V. Analysis of the final data set

Citation
Pg. Martin et al., Ultraviolet interstellar linear polarization. V. Analysis of the final data set, ASTROPHYS J, 510(2), 1999, pp. 905-914
Citations number
37
Language
INGLESE
art.tipo
Article
Categorie Soggetti
Space Sciences
Journal title
ASTROPHYSICAL JOURNAL
ISSN journal
0004-637X → ACNP
Volume
510
Issue
2
Year of publication
1999
Part
1
Pages
905 - 914
Database
ISI
SICI code
0004-637X(19990110)510:2<905:UILPVA>2.0.ZU;2-W
Abstract
Using recent measurements of ultraviolet interstellar polarization, we have examined its relationship to ultraviolet extinction and to polarization an d extinction measurements in the visible and infrared. The relationship bet ween the relative amount of ultraviolet polarization and the parameter lamb da(max), determined using only visible data, is confirmed and strengthened, for example, by a tight correlation between p(6 mu m(-1))/P-max and lambda (max)(-1). A good fit to the wavelength dependence of the polarization from the infrared to the ultraviolet can be achieved with a five-parameter func tion combining a power law in the infrared and a Serkowski-like function in the ultraviolet. The polarization efficiency (ratio of polarization to ext inction) is less in the ultraviolet than in the visual, and the ratio of th ese efficiencies increases systematically with lambda(max)(-1). We relate t hese effects to systematic changes in the underlying aligned grain size dis tribution. The polarization efficiency of the grains causing the 2175 Angst rom extinction bump along most sight lines is so (unusually) small that no statistically significant polarization feature is detectable in that wavele ngth region. Only two of 28 sight lines show a definite polarization featur e. The environments of the two bump sight lines are similar but not unique, and the mechanism for producing the polarization feature along only these two sight lines is still not known.