Exploring and modeling high-excitation emission in the ejecta and the Wind of Eta Carinae.

Abstract

Eta Carinae (! Car) is the most massive, most luminous star in our region of the Galaxy. It is an evolved massive star system and therefore provides many clues to the fate of the most massive stars. In the 1840s its unstable nature culminated in the Great Eruption when it briefly became the second brightest star in the sky and ejected more than ten solar masses, which today enshroud the surviving star as a bipolar nebula. The “supernova impostor” phenomenon and its aftermath constitute a major gap in the theory of massive stars, and ! Car is the only example that can be studied in detail. Its recovery has been unsteady with unexplained photometric and spectral changes in the 1890s and 1940s. Combining data from HST STIS and Gemini-S GMOS between 1998 and 2010, I analyzed several spectroscopic cycles that occur every 5.54 years. In addition, I used some data from the VLT UVES, Magellan II MIKE, and Ir´en´ee du Pont B&C instruments. Observations with a variety of different slit position angles made it possible to map the emission across the nebula and the complex outer ejecta of ! Car permit to observe the star at different stellar latitudes via reflected light. In order to study the distribution of gas and ionizing radiation around ! Car and their implications for its likely companion star, I examined several high-excitation emission lines. The principal results are: (1) The high-excitation fluxes varied systematically and non-trivially throughout ! Car’s 5.5-year spectroscopic cycle. (2) A brief, strong secondary maximum occurred just before the 2003.5 spectroscopic event. (3) These emission lines are strongly concentrated at the Weigelt knots several hundred AU northwest of the star. With less certainty, [Ne III] appears to be somewhat more concentrated than [Fe III]. (4) A faster, blueshifted component appears concentrated near the star and elongated perpendicular to the system’s bipolar axis. (5) Using the photoionization program Cloudy, I estimated the range of parameters for the hot secondary star that would give satisfactory high-excitation line ratios in the close ejecta; Teff ! 40, 000 K, L " 4×105 L!, and Minit " 40–50 M!, for example, would be satisfactory. The allowed region in parameter space is wider and mostly less luminous than some previous authors suggested. Spectra obtained with Gemini GMOS throughout 2007–2010 were used to observe the 2009 spectroscopic event from different stellar latitudes. The He II "4687 emission, only observed during the “events,” was analyzed in spectra in direct view and in reflected polar-on spectra at FOS4. The time-delay of He II at FOS4 is about 18 days and therefore consistent with the predicted additional time-travel time at FOS4. The equivalent width and radial velocity behavior of He II at FOS4 mirrors the behavior observed in direct view. These findings imply a symmetric geometry for the origin of the He II emission and are difficult to reconcile with some proposed He II emitting regions and some orbital models. H I, He I, and Fe II lines, observed at different latitudes, reveal details about the changing wind during the “event.” N II ""5668–5712 emission and absorption lines behave qualitatively like the He I lines. Spectral lines of ! Car’s stellar wind regions can be classified into four physically distinct categories: 1) low-excitation emission such as H I and Fe II, 2) higher excitation He I features, 3) N II lines, and 4) He II emission. These categories have different combinations of radial velocity behavior, excitation processes, and dependences on the secondary star. In this sense the N II features resemble the He I lines, but they represent zones of lower ionization. This combination of attributes appears to be unique in ! Car’s well-observed spectrum. N II probably excludes some proposed models, such as those where He I lines originate in the secondary star’s wind or in an accretion disk. Spectra in 2009 and 2010 showed that major stellar-wind emission features in the spectrum of ! Car have recently decreased by factors of order 2 relative to the continuum. This is unprecedented in the modern observational record. The simplest explanation is a rapid decrease in the wind density. Work presented in this thesis was published in Mehner et al. (2010a,b, 2011).

v