Submillimeter and infrared studies of mass lost by asymptotic giant branch stars

Citation

Young, Kenneth Harbour (1994) Submillimeter and infrared studies of mass lost by asymptotic giant branch stars. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/pkm9-jk53. https://resolver.caltech.edu/CaltechETD:etd-09172008-085430

Abstract

This thesis consists of four separate studies of the material ejected from highly evolved red giant stars, preceded by a brief introductory chapter.

Chapter 2 presents the results of an attempt to identify all the post main sequence stars in the sky with circumstellar shells large enough to have been resolved by the IRAS satellite. Both the 60μm and 100μ survey mode data were used. A total of 512 stars were examined, including all evolved stars with envelopes which had been detected in surveys of mm or submm CO emission. A total of 76 stars were found to have shells with radii exceeding 2' in the 60μ data. Most of these resolved shells surround semiregular variable stars. Although only 40% of the resolved objects are carbon stars, the fraction of carbon stars with resolved envelopes is higher than that of oxygen—rich stars. The linear sizes of these shells were calculated; the average shell size is 0.76 pc. If a measurement of the expansion velocity was available from CO observations, the age of the shell was estimated using a simple model of the shell's interaction with the ISM. Some of the shells appear to be detached, and in most cases shells with inner radii greater than ~ 0.3 pc have no detectable CO. The duration of mass loss for Mira variables and carbon stars is estimated to be 10[superscript 5] and 2 x 10[superscript 5] years, respectively. The procedure by which the data were analysed is explained in the appendix to chapter 2, which also contains the results of several tests performed to verify the results statistically.

A survey of CO(3-2) emission from Mira variables comprises chapter 3. The survey examined nearby oxygen—rich Miras, without regard to their infrared properties. Several new detections were made, including some circumstellar shells with expansion velocities lower than any previously reported. Mass loss rates for the detected stars were calculated. While the mass loss rate shows significant correlation with all 4 IRAS fluxes (when normalized by distance), it appears completely uncorrelated with any color derived from these fluxes. A star's spectral type at maximum light was found to be a good predictor of which stars would be detectable; no star earlier than M5.5 was detected even though some such stars had large IRAS fluxes. A power—law relationship was found between the envelope expansion velocity and the star's mass loss rate. Extrapolation of this result implies that envelopes with expansion velocities above ~ 18 km s[superscript -1] must be optically thick.

The peculiar submm emission line profiles of CRL 2688, a protoplanetary nebula, are examined in chapter 4. The molecular envelope was found to have three kinematic components, including a previously undetected ~ 100 km s[superscript -1] wind. The mechanical momentum in this high velocity wind appears to be much too large to have been supplied by radiation pressure. A strong self-absorption feature seen in the CO emission lines was examined at very high spectral resolution, and was found to have an extremely sharp blue edge. The sharpness of this feature shows that the velocity field in the extended envelope is very well ordered.

The final chapter examines atomic carbon emission in the envelope of the young planetary nebula NGC 7027. Neutral carbon, traced by the 609μm hyperfine line, is extended over a region 30-40" in diameter, and appears to be coextensive with CO throughout most of the neutral envelope. A 158μm spectrum of C II was used to calculate a lower limit to the C II mass. There appears to be roughly equal amounts of CO, C I and C II in the inner neutral envelope. Most of the atomic carbon must have been liberated by photodissociation of CO. Estimates of the mass loss rate for this object derived from CO observations may be too low by a factor of ~ 3.

Thesis Files