A quasi-time-dependent radiative transfer model of OH 104.9+2.4

¹ Max-Planck-Institut für Astronomie, Königstuhl 17, 69117 Heidelberg, Germany e-mail: riechers@mpia-hd.mpg.de
² Max-Planck-Institut für Radioastronomie, Auf dem Hügel 69, 53121 Bonn, Germany
³ Special Astrophysical Observatory, Nizhnij Arkhyz, Zelenchuk district, Karachai-Cherkessian Republic, Russia
⁴ Institute for Computational Astrophysics, Saint Mary's University, Halifax, NS B3H 3C3, Canada
⁵ Crimean Astrophysical Observatory, Nauchny, 98409, Crimea, Ukraine

Received: 7 October 2004
Accepted: 8 February 2005

Abstract

We investigate the pulsation-phase dependent properties of the circumstellar dust shell (CDS) of the OH/IR star OH 104.9+2.4 based on radiative transfer modeling (RTM) using the code DUSTY. Our previous study concerning simultaneous modeling of the spectral energy distribution (SED) and near-infrared (NIR) visibilities (Riechers et al. 2004) has now been extended by means of a more detailed analysis of the pulsation-phase dependence of the model parameters of OH 104.9+2.4. In order to investigate the temporal variation in the spatial structure of the CDS, additional NIR speckle interferometric observations in the K' band were carried out with the 6 m telescope of the Special Astrophysical Observatory (SAO). At a wavelength of m the diffraction-limited resolution of 74 mas was attained. Several key parameters of our previous best-fitting model had to be adjusted in order to be consistent with the newly extended amount of observational data. It was found that a simple rescaling of the bolometric flux F_bol is not sufficient to take the variability of the source into account, as the change in optical depth τ over a full pulsation cycle is rather high. On the other hand, the impact of a change in effective temperature T_eff on SED and visibility is rather small. However, observations, as well as models for other AGB stars, show the necessity of including a variation of T_eff with pulsation phase in the radiative transfer models. Therefore, our new best-fitting model accounts for these changes.