2009A&A...501..999P

A strong signature of a circumstellar disc around a high-mass protostar has been inferred from high resolution methanol maser observations in NGC7538-IRS1 N. This interpretation has however been challenged, with a bipolar outflow proposed as an alternative explanation. We compare the two proposed scenarios for best consistency with the observations. Using a newly developed formalism, we model the optical depth of the maser emission at each observed point in the map and LOS velocity for the two scenarios. We find that if the emission is symmetric around a central peak in both space and LOS velocity, then it has to arise from an edge-on disc with sufficiently fast differential rotation. Disc models successfully fit ∼100 independent measurement points in position-velocity space with 4 free parameters to an overall accuracy of 3-4%. Solutions for Keplerian rotation require a central mass of at least 4M_☉. Close to best-fitting models are obtained if Keplerian motion is assumed around a central mass equaling ∼30 M_☉, as inferred from other observations. In contrast, we find that classical bipolar outflow models cannot fit the data, although it could be applicable in other sources. Our results strongly favour the differentially rotating disc hypothesis to describe the main feature of the 12.2 (and 6.7)GHz methanol maser emission in NGC7538 IRS1 N. Furthermore, for Keplerian rotation around a ∼30M_☉ protostar, we predict the position and velocity at which tangentially amplified masers should be detected in high dynamic range observations. Also, our model predicts the amplitude of the proper motion of some of the maser features in our data. Confirmation of a large central mass would strongly support the idea that even the highest mass stars (>20M_☉) form via accretion discs, similarly to low-mass stars. Finally we note that our new formalism can readily be used to distinguish between discs and outflows for thermal emitting line sources as well as masers.