Search

To advance our understanding of the relationship between stellar chromospheres and mass loss, which is a common property of carbon stars and other asymptotic giant branch stars, we have obtained ultraviolet spectra of the nearby N-type carbon star UU Aur using the Hubble Space Telescope (HST). In this paper we describe the HST observations, identify spectral features in both absorption and emission, and attempt to infer the velocity field in the chromosphere, upper troposphere, and circumstellar envelope from spectral line shifts. A mechanism for producing fluoresced emission to explain a previously unobserved emission line is proposed. Some related ground-based observations are also described.

The outer layers and near circumstellar envelope (CSE) of a typical carbon star have been studied using available data from theoretical and empirical models. An attempt is made to match the density-velocity structure of the photosphere-chromosphere region to values from the radio CO observations, which arise from the outer CSE. It is concluded that the stellar atmosphere includes a relatively thin high-temperature region close to hydrostatic equilibrium and a much more extended cooler region of outflowing gas and dust. To extend the outer photosphere and chromosphere to match the mass loss density appears to require an injection of energy and momentum by some mechanism rather close to the stellar surface.

A millimeter-wave survey has been made of bright relatively unobscured, carbon stars, chosen on the basis of their optical properties. Out of 26 program objects, (J = 2-1)CO emission is detected from 15. Most of these had not been previously detected. There are many differences among the observed objects, but one rather interesting trend emerges: a positive correlation, at moderate IR excesses, between the IR dust emission and the expansion velocity of the dense wind. A similar, positive correlation with the mass-loss rate implies that stars with larger mass fluxes also accelerate them to larger velocities. At high-IR excesses, both correlations break down, and the momentum rate may be limited by the momentum rate of the stellar radiation. All these effects could be ascribed to differences in the gas-to-dust ratio, assuming that radiation pressure initiates and accelerates the wind.

A binary model for the carbon star HD 59643 is discussed in which the secondary spectrum is formed in an accretion disk. If this hot, ultraviolet-emitting disk radiates like a 20,000 K black-body, it must be 0.03 solar radii or less across at minimum emission. Large widths of C IV multiplet UV1 on high-resolution spectra indicate its formation in the inner parts of a disk. The semiforbidden C III and Si III lines, however, are much narrower and could be formed in the outer parts of a disk or in the carbon star's chromosphere. The electron density in the region of formation of C III is about 10 to the 10th/cu cm.

A high-resolution LWP IUE spectrum of the bright N-type carbon star TX Psc demonstrates that the Mg II h and k emission profiles are strongly affected by absorption from Mg II, Mn I, probably Fe I, and possibly from molecules. The indication that the absorbing matter has a column density of not less than 10 to the 20th H atoms or molecules per sq cm is consistent with absorption in a slowly expanding envelope. The integrated Mg II line flux is found to be much greater than in 1981, and the radio CO (J = 1 - 0) line from the circumstellar shell is detected. Results for a column density of not larger than 10 to the 22nd H2 molecules/sq cm, and a radial velocity close to that of the star, are in agreement with those obtained from UV data. Some dust emission from carbon grains is suggested by the far infrared flux distribution, and a mass-loss rate estimation for the star of 10 to the -6th to 10 to the -8th solar masses is obtained.

In the present paper, the observation of a white dwarf companion of the mild Ba star Xi-1 Ceti (= 65 Ceti = HR 649 = HD 13611) is reported, taking into account also the properties of the mild Ba star and of its companion. The UV spectrum of Xi-1 Ceti is discussed along with an interpretation of this spectrum. Attention is given to the effective temperature of the companion, the absorption bands in the spectrum, the radius and mass of the Xi-1 Ceti companion, and questions regarding the obscuration of the companion by the atmosphere of the Ba star. It is found that the overall energy distribution of the Xi-1 Ceti companion can best be matched with a 14,000 K DA white dwarf of log g = 8 or less. However, the absolute intensity is too small and would require a radius too small and a mass too large for such a gravity.

Motivated by recent work which hints at a possible deficiency of hydrogen in non-Mira N-type carbon stars and to further explore the parameter space of chemical composition, computations have been made of a series of hydrogen-deficient models for carbon stars. For these models Teff = 3000 K, and log g = 0.0. Solar abundances are used for all elements except for carbon (which is enhanced to give C/O = 1.05), hydrogen, and helium. As the fractional abundance of hydrogen is decreased, being replaced by helium, the temperature-optical depth relation is affected only slightly, but the temperature-pressure relation is changed. The most striking change in the emergent flux is the decrease of the H(-) peak at 1.65 micron compared with the blackbody peak at 1.00 micron.

Low-resolution IUE spectra of 23 cool giant stars ranging from K3 through M6 have been used to examine chromospheric activity in late-type giants. Previous studies have suggested a decrease in the fractional flux, normalized to bolometric flux, of the Mg II resonance lines (an indicator of chromospheric activity) with effective temperature. Using more accurately determined effective temperatures, this decrease is confirmed and found to be more rapid than previously suggested. A strong correlation is found between relative TiO band strengths, as measured by the Wing TiO index, and the level of chromospheric activity, as measured by Mg II fluxes. Cool giants which have weak TiO bands relative to the mean TiO strengths for stars of similar color also have Mg II lines that are weaker than the mean. Changes in composition and in chromospheric activity as a consequence of aging are suggested as a possible origin of this interesting relation.

Long-wavelength IUE spectra of 13 normal R stars and two hydrogen-deficient R0 supergiants were obtained. Early R stars are noted to have line spectra and levels of flux in the ultraviolet characteristic of G5-K2 III stars, whereas late R stars were observed to have colors and line spectra similar to late K and M stars, but with greatly enhanced strength of low-lying multiplets of neutral metals. Hydrogen-deficient carbon stars show readily apparent differences from the normal early R stars, reflecting their luminosity and somewhat higher temperatures. The lines of neutral metals in these stars are weakened, while those of ionized metals are strengthened.

Ultraviolet IUE spectra of N, R, and S stars are displayed and discussed. Carbon stars of type N have IUE spectra similar to M stars except that the Mg II lines are much weaker and low-excitation lines from neutral metals are stronger. Spectra of early R stars (R0-R3) closely resemble those of late G or early K in both lines and continuum with no significant emission lines. Late R stars (R5-R8) show spectra like late K or early M stars except that lines from low levels of excitation are stronger. Ultraviolet spectra of S stars show a wide variation in the strength of Mg II lines.

For galactic N and late R type irregularly variable carbon stars, a comparison is made of airborne infrared spectrophotometry to spectral flux distributions derived from model atmospheres. Although the agreement is good in several respects, the H(-) emission peak at 1.6 microns, so prominent in the models, is significantly weakened in the stellar spectrophotometry. It seems unlikely that molecular bands in the stars can completely obliterate this peak. A plausible explanation is that hydrogen is deficient in these stars, most probably from an episode of mass loss. The resultant cold shells surrounding these stars should be amenable to observation by IRAS, or other far-infrared observatories.

Identifications for absorption features detected in the ultraviolet spectrum of the hydrogen-deficient supergiant carbon star HD 182040 observed at high resolution with the IUE satellite are presented. The spectrum in this range is filled with blended lines of the iron-peak elements, with Fe I and Cr II dominating the spectrum. Spectra definitely and probably detected are given, as well as those searched for and not detected. In the investigated range, the spectrum resembles that of a middle-G supergiant. The absorption-line spectrum is similar to the spectra of the sun and other late-type stars, but neutral metal lines are weaker than in K and M giants. The level of ionization is significantly higher than in the sun, Zeta Cap, and Epsilon Gem.

Current/PAD research in the area of distributed data base management considers facilities for supporting CAD/CAM data management in a heterogeneous network of computers encompassing multiple data base managers supporting a variety of data models. These facilities include coordinated execution of multiple DBMSs to provide for administration of and access to data distributed across them.

General capabilities of data base management technology are described. Information requirements posed by the space station life cycle are discussed, and it is asserted that data base management technology supporting engineering/manufacturing in a heterogeneous hardware/data base management system environment should be applied to meeting these requirements. Today's commercial systems do not satisfy all of these requirements. The features of an R&D data base management system being developed to investigate data base management in the engineering/manufacturing environment are discussed. Features of this system represent only a partial solution to space station requirements. Areas where this system should be extended to meet full space station information management requirements are discussed.

Spectra of several carbon stars in the two micron region are examined and compared with synthetic spectra from model atmospheres. Molecular hydrogen has been directly observed in carbon stars for the first time through the quadrupole 1-0 S(1) line at 4712.9/cm. It is definitely present in several Mira variables, and is definitely not present in several irregular variables and one Mira variable. A discrepancy between the predicted and observed line depths at different temperatures is shown. Resolution of the discrepancy by change of temperature alone would require an upward revision of the present effective temperature scale by at least 600 K. Several possible resolutions of the dilemma are briefly examined.

The paper reports the successful detection of the ultraviolet spectra of three N-type carbon stars: BL Ori, TX Psc, and T Ind. Exposures of approximately 2-3 hours with the LWR camera of the IUE Satellite at low dispersion were necessary to record the ultraviolet flux. The shortward extension of the photospheric continuum is well exposed down to about 2800 A, with strong molecular and atomic absorption features. The spectrum appears not to be seriously affected by the unknown opacity source which blankets N stars (especially redder ones) in the violet. Emission lines of Mg II (h and k) and C II (2325 A) are conspicuous, indicating the presence of chromospheres in the stars. The ratios of integrated fluxes in the Mg II lines and C II lines to bolometric flux are respectively much smaller and slightly smaller than values observed in oxygen-rich M giants and supergiants.

Low dispersion long wave radiation spectra of three early N-type carbon stars (BL Orionis, TX Piscium and T Indi) obtained with the IUE are discussed. N-star spectra were compared with IUE observations of late M-type stars and several differences were noted. The Mg II and C II chromospheric emission lines reported are the first convincing detection of a 10,000-20,000 K chromosphere in the N stars.

An intercomparison of model stellar atmospheres and observations of real stars can lead to a better understanding of the relationship between the physical properties of stars and their observed radiative flux. In this spirit we have determined wide-band and narrow-band magnitudes and colors for a subset of models of K and M giant and supergiant stars selected from the grid of 40 models by Johnson, Bernat and Krupp (1980) (hereafter referred to as JBK). The 24 models selected have effective temperatures of 4000, 3800, 3600, 3400, 3200, 3000, 2750 and 2500 K and log g = 0, 1 or 2. Emergent energy fluxes (erg/ sq cm s A) were calculated at 9140 wavelengths for each model. These computed flux curves were folded through the transmission functions of Wing's 8-color system (Wing, 1971; White and Wing, 1978) and through Johnson's (1965) wide-band (BVRIJKLM) system. The calibration of the resultant magnitudes was made by using the absolute calibration of the flux curve of Vega by Schild, et al. (1971).

An overview of the capabilities and software architecture of the IPAD information processor (IPIP) is presented. IPIP is a state-of-the-art data base management system that satisfies engineering requirements not addressed by present day commercial systems. It also significantly advances a number of capabilities that are offered commercially. IPIP capabilities range from support for multiple schemas and data models to support for distributed processing, configuration control, and data inventory management. IPIP exploits semantic commonality in features offered in various forms at different user interfaces in today's commercial systems. An integrated software architecture supports all user interfaces: programming languages, interactive data manipulation, and schema languages. This approach promotes simplicity and compactness in software and permits features to be offered symmetrically across all appropriate user interfaces.

A method for treating atomic and molecular line opacities in cool stellar atmospheres by a statistical opacity sampling is investigated. Under the usual assumptions of plane-parallel geometry, radiative equilibrium, hydrostatic equilibrium, and LTE, each radiative quantity is computed monochromatically at each chosen frequency and depth without any averaging of the opacity. The number of frequencies needed to allow an accurate integration of the energy flux over a given spectral interval is investigated as a function of depth, including opacity for both CN and C2. This method is extended to the calculation of a model atmosphere of a star, and the effect of the number and placement of frequency points is studied. The method is applied to treating molecular lines of CO, C2, and CN in a cool carbon star. Significant advantages of the opacity sampling method are its flexibility, which permits computation of models having arbitrary variations of chemical composition and of opacity with wavelength and depth, and generalizability to include departures from LTE.

Molecular column densities are presented for 35 molecules in a variety of cool stellar model atmospheres. From an examination of the predicted column densities, we draw the following conclusions: (1) OH might be visible in carbon stars which have been generated from triplet-alpha burning, but will be absent from carbon stars generated from the CNO bi-cycle; (2) the TiO/ZrO ratio shows small but interesting variations as C/O is changed and as the effective temperature is changed; (3) the column density of silicon dicarbide (SiC2) is sensitive to abundance, temperature, and gravity; hence, all relationships between the strength of SiC2 and other stellar parameters will show appreciable scatter. There is, however, a substantial luminosity effect present in the SiC2 column densities; (4) unexpectedly, SiC2 is anticorrelated with C2; (5) the presence of SiC2 in a carbon star allows us to eliminate the possibility that these stars are both 'hot' (T sub eff greater than or equal to 3000 K) and have been produced through the CNO bi-cycle (so that C/H is less than solar).

An opacity sampling (OS) technique for treating the radiative opacity of large numbers of atomic and molecular lines in cool stellar atmospheres is presented. Tests were conducted and results show that the structure of atmospheric models is accurately fixed by the use of 1000 frequency points, and 500 frequency points is often adequate. The effects of atomic and molecular lines are separately studied. A test model computed by using the OS method agrees very well with a model having identical atmospheric parameters computed by the giant line (opacity distribution function) method.