ABSTRACT
AIMS: Extragalactic observations of water emission can provide valuable insights into the excitation of the interstellar medium. In particular they allow us to investigate the excitation mechanisms in obscured nuclei, i.e. whether an active galactic nucleus or a starburst dominate. METHODS: We use sub-arcsecond resolution observations to tackle the nature of the water emission in Arp 220. ALMA Band 5 science verification observations of the 183 GHz H2O 313-220 line, in conjunction with new ALMA Band 7 H2O 515-422 data at 325 GHz, and supplementary 22 GHz H2O 616 - 523 VLA observations, are used to better constrain the parameter space in the excitation modelling of the water lines. RESULTS: We detect 183 GHz H2O and 325 GHz water emission towards the two compact nuclei at the center of Arp 220, being brighter in Arp 220 West. The emission at these two frequencies is compared to previous single-dish data and does not show evidence of variability. The 183 and 325 GHz lines show similar spectra and kinematics, but the 22 GHz profile is significantly different in both nuclei due to a blend with an NH3 absorption line. CONCLUSIONS: Our findings suggest that the most likely scenario to cause the observed water emission in Arp 220 is a large number of independent masers originating from numerous star-forming regions.
ABSTRACT
CONTEXT: Recent observations at subarcsecond resolution, now possible also at submillimeter wavelengths, have shown intricate circumstellar structures around asymptotic giant branch (AGB) stars, mostly attributed to binary interaction. The results presented here are part of a larger project aimed at investigating the effects of a binary companion on the morphology of circumstellar envelopes (CSEs) of AGB stars. AIMS: AGB stars are characterized by intense stellar winds that build CSEs around the stars. Here, the CO(J = 3â2) emission from the CSE of the binary S-type AGB star W Aql has been observed at subarcsecond resolution using ALMA. The aim of this paper is to investigate the wind properties of the AGB star and to analyse how the known companion has shaped the CSE. METHODS: The average mass-loss rate during the creation of the detected CSE is estimated through modelling, using the ALMA brightness distribution and previously published single-dish measurements as observational constraints. The ALMA observations are presented and compared to the results from a 3D smoothed particle hydrodynamics (SPH) binary interaction model with the same properties as the W Aql system and with two different orbital eccentricities. Three-dimensional radiative transfer modelling is performed and the response of the interferometer is modelled and discussed. RESULTS: The estimated average mass-loss rate of W Aql is M = 3.0×10-6 Mâ yr-1 and agrees with previous results based on single-dish CO line emission observations. The size of the emitting region is consistent with photodissociation models. The inner 10â³ of the CSE is asymmetric with arc-like structures at separations of 2-3â³ scattered across the denser sections. Further out, weaker spiral structures at greater separations are found, but this is at the limit of the sensitivity and field of view of the ALMA observations. CONCLUSIONS: The CO(J = 3â2) emission is dominated by a smooth component overlayed with two weak arc patterns with different separations. The larger pattern is predicted by the binary interaction model with separations of ~10â³ and therefore likely due to the known companion. It is consistent with a binary orbit with low eccentricity. The smaller separation pattern is asymmetric and coincides with the dust distribution, but the separation timescale (200 yrs) is not consistent with any known process of the system. The separation of the known companions of the system is large enough to not have a very strong effect on the circumstellar morphology. The density contrast across the envelope of a binary with an even larger separation will not be easily detectable, even with ALMA, unless the orbit is strongly asymmetric or the AGB star has a much larger mass-loss rate.
