CO line emission from circumstellar envelopes

¹ Departamento de Astrofísica Molecular e Infrarroja, IEM, CSIC, 28006 Madrid, Spain
² Space Research Organization Netherlands, P.O. Box 800, 9700 AV Groningen, The Netherlands e-mail: teyssier@sron.rug.nl
³ Observatorio Astronómico Nacional, Apartado 1143, 28800 Alcalá de Henares, Spain e-mail: v.bujarrabal@oan.es
⁴ Caltech Submillimeter Observatory, 111 Nowelo Street, Hilo, HI 96720, USA e-mail: hiro@submm.caltech.edu
⁵ California Institute of Technology, Pasadena, CA 91125, USA e-mail: phillips@submm.caltech.edu

Received: 4 July 2005
Accepted: 30 December 2005

Abstract

Aims.We present the results of a multi-transition CO observational program conducted on a sample of AGB and post-AGB stars envelopes. We have collected maps and single pointing observations of these envelopes in 5 rotational transitions ranging from J = 1–0 to J = 6–5, including in particular new observations of the CO line at 691 GHz at the CSO. The use of such a set of mm and submm CO line on stellar envelopes is rare and limited to the work of some authors on IRC+10216.Methods.Using a model for the CO emission of an AGB circumstellar envelope, in combination with a standard LVG approach, we have conducted a systematic modelling analysis using the whole set of CO data collected for a sample of 12 sources. We simultaneously fit all five transitions, taking into account the spatial information provided by the maps.Results.We find mass-loss rates in the range 1  10^-7 to 4  10/yr, and envelope temperatures ranging from 20 K to 1000 K at a radius of 10¹⁶ cm. There seem to be a general anti-correlation between mass loss rates and temperature, the high mass loss rate AGBs having low temperatures, and vice versa. We show that most AGB data can be fitted using a constant mass loss rate, at least within the calibration uncertainties associated with the data collected at different frequencies. For some cases though (e.g. CIT 6, R Hya, χ Cyg), a change in the mass loss rate history needs to be invoked to reconcile data at low- and high-J, a scenario already mentioned by several authors to explain observations of WX Psc.