Mass loss and rotational CO emission from Asymptotic Giant Branch stars

¹ Astronomical Institute “Anton Pannekoek”, University of Amsterdam, Kruislaan 403, 1098 SJ Amsterdam, The Netherlands
² UCLA, Division of Astronomy, 405 Hilgard avenue, Los Angeles, CA 90095-1562, USA
³ Max-Planck-Institut für Radioastronomie, Auf dem Hügel 69, 53121 Bonn, Germany
⁴ Stockholm Observatory, SCFAB, 106 91 Stockholm, Sweden
⁵ Kapteijn Institute, University of Groningen, PO Box 800, 9700 AV Groningen, The Netherlands
⁶ SRON Laboratory for Space Research, PO Box 800, 9700 AV Groningen, The Netherlands
⁷ Instituut voor Sterrenkunde, Katholieke Universiteit Leuven, Celestijnenlaan 200B, 3001 Heverlee, Belgium
⁸ NASA Ames Research Center, Mail Stop 245-6, Moffett Field, CA 94035-1000, USA

Corresponding author: F. Kemper, kemper@astro.ucla.edu

Received: 2 January 2003
Accepted: 6 May 2003

Abstract

We present submillimeter observations of rotational transitions of carbon monoxide from up to for a sample of Asymptotic Giant Branch stars and red supergiants. It is the first time that the high transitions and are included in such a study. With line radiative transfer calculations, we aim to determine the mass-loss history of these stars by fitting the CO line intensities. We find that the observed line intensities of the high transitions, including the transition, are significantly lower than the predicted values. We conclude that the physical structure of the outflow of Asymptotic Giant Branch stars is more complex than previously thought. In order to understand the observed line intensities and profiles, a physical structure with a variable mass-loss rate and/or a gradient in stochastic gas velocity is required. A case study of the AGB star WX Psc is performed. We find that the CO line strengths may be explained by variations in mass-loss on time scales similar to those observed in the separated arc-like structures observed around post-AGB stars. In addition, a gradient in the stochastic velocity may play a role. Until this has been sorted out fully, any mass loss determinations based upon single CO lines will remain suspect.