8 Summary

We set out to simulate H₂O maser emission in the circumstellar environment of a M-Mira variable, compare our results qualitatively with observations and discuss them in terms of the physical conditions prevailing in the CE. The combination of an H₂O maser saturation radiation transport model with a M-Mira CE pulsation model has resulted in:

1.

Synthetic single-dish lineshapes at 22, 183, 321 and 325GHz which have velocity extents consistent with those observed. Singly-peaked profiles are reproduced for 22, 325 and 183GHz maser emission, as observed for stars of similar ${\dot{M}}$ to our model star. The ratio of the peak synthetic lineshape intensities is comparable to those observed.

2.

A synthetic 22GHz image which has an asymmetric distribution of bright components within a spatial extent typical of observations towards Miras. Our simulations confirm that bright 22GHz emission originates from a zone in which material is undergoing acceleration away from the star, as observed.

3.

A prediction of the spatial structure of 183, 325 and 321GHz masers in the CE. We find that these masers occupy overlapping regions in the CE, with the 321GHz masers originating from the same region as SiO masers. We attribute the observed variability characteristics of these masers to their very different extents in the CE, and the difference in the number of bright components contributing to the total emission for each transition.

4.

A prediction of the existence of new, bright stellar masers. The lineshapes and spatial structure of these masers in the CE have been calculated, on the basis of which we are also able to predict their variability. Future experiments using ALMA and the HSO may allow verification of these predictions.

An important result of these simulations is that bright maser emission often occurs from physical conditions which have not been explored by previous H₂O maser models. These aimed to find the optimum conditions for H₂O maser emission, rather than input conditions from advanced stellar models. We find here that bright maser emission can occur from hotter regions than those previously considered, of $T_{\rm k}> 2000$ K, although we note that the presence of higher temperatures in the H₂O maser zone is uncertain. In addition, propagation of maser emission including velocity-shifts along the line-of-sight plays a crucial role in deciding which maser components in the CE will achieve very high intensities.

Finally, it is striking that these simulations reproduce both the stellar SiO (see H96) and H₂O maser properties. This numerical experiment shows that a mechanism of pulsation and dust radiation pressure for driving AGB mass loss is qualitatively consistent with observed stellar SiO and H₂O maser data.

Acknowledgements

We thank the Swedish Foundation for International Cooperation in Research and Higher Education (STINT) for the financial support of E. M. L. H. This work was carried out on the Miracle supercomputer, at the HiPerSPACE Computing Centre, UCL, which is funded by the UK Particle Physics and Astronomy Research Council. We thank the referee Dr. M. Reid for his significant contributions to this paper.