No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Epsilon Aurigae. II - The shell spectrum
Abstract
A refined mathematical procedure is applied to obtain the shell spectrum
produced by the gaseous envelope of the eclipsing companion of Epsilon
Aurigae during the 1982 to 1984 eclipse. The extraction algorithm
compensates for the spectrum of the primary FO supergiant and also for
its intrinsic variability that is translated into a residual uncertainty
band in a fully quantitative manner. Fifteen pure shell spectra were
derived at different eclipse phases, covering the range from 3600A to
4960A and providing a complete scanning through the gaseous body,
including its outer edge well beyond the eclipsing body.
The mysterious star system epsilon Aurigae undergoes an eclipse every 27.1 years that lasts nearly two years. The most recent eclipse started during the late summer of 2009. An international campaign for observing this eclipse was created in 2006, with a web site for information and, to-date, 17 periodic newsletters for details, as well as a Yahoo forum List for immediate announcements and comments. Photometric data in the UBVRIJH bands have been submitted. Ingress occurred with first contact in the V band estimated at the second week of 2009 August and second contact estimated at 2010 mid-January. Spectroscopic data were also obtained during ingress. Spectroscopic data have been provided in the potassium I region, hydrogen alpha and beta regions and sodium D line region of the star system's spectrum. In this paper we describe details of observations and preliminary analysis during ingress and second contact. We introduce the observers and discuss plans for observing throughout totality and the end of the eclipse in 2011.
A sample of high-dispersion optical spectrograms of ε Aur, taken on different epochs before and during eclipse, has been processed by computer in order to extract the shell spectrum, produced by the eclipsing body. Using an uneclipsed spectrum as reference, the primary's dominant contribution was removed, and this makes it possible for the first time to study the secondary body spectrum separately at all eclipse phases. As the first striking result, one finds that the internal regions of the extended eclipsing body are hotter than the outer zones.
The radial velocities and equivalent widths of absorption lines due to
the gas in the secondary component of Epsilon Aur are measured at
wavelengths 3700-7700 A on the spectrograms taken during the 1982-1984
eclipse. In the extended part of the secondary, hydrogen is distributed
like a disk, and metallic densities decrease in the inner part of the
disk. The excitation temperature of metals is about 4000 K, but some hot
gas exists in the ingress-side disk. Lyman photons of the primary
produce a shock wave in the ingress-side disk and decelerate the
rotating gas by about 10 km/s. THe H-alpha profiles and the radial
velocities of metallic lines fit a Keplerian disk model of the secondary
which is rotating with velocity of 42 + or - 2 km/s at the egress-side
limb. The rotation curve rejects the binary model consisting of massive
components usually accepted so far. For an astrometric distance of 580
pc, a model with the primary mass of 1.5-2.5 solar masses is consistent
with the luminosities of the components and the projected area of the
secondary disk derived from the optical and infrared photometry. The
primary is almost on the Eddington limit.
Optical high-resolution 2.7-m McDonald Observatory observations of Epsilon Aur obtained during the 1982-1985 eclipse are discussed, with focus on the K I resonance lines, some Fe II and Ti II lines in the blue, and the near-infrared N I excitation lines. The N I lines are immune to contamination by the secondary, and their radial velocity decline over the three-year interval fits the predicted velocity curve. The 7664 A and 7699 A K I lines strengthen greatly during the eclipse due to gas in orbit around the secondary passing in front of the primary, and the velocity pattern analysis indicates a mass of the secondary to solar mass ratio of between about 3 and 6 and a primary mass of 3 solar masses. The data are consistent with the picture of a low-mass post-AGB primary having lost mass through a wind and by Roche-lobe transfer to the secondary, and which will evolve to a white dwarf in less than 10,000 years.