Clumpy ultracompact H II regions. III. Cometary morphologies around stationary stars.

ABSTRACT Cometary ultracompact H II regions have been modelled as the interaction of thehypersonicwindfromamovingstarwiththemolecularcloudwhichsurroundsthestar.We here show that a similar morphology can ensue even if the star is stationary withrespect to the cloud material. We assume that the H II region is within a stellar windbubble which is strongly mass loaded: the cometary shape results from a gradient inthe distribution of mass loading sources. This model circumvents problems associatedwith the necessarily high spatial velocities of stars in the moving star models.Key words: hydrodynamics – stars: mass-loss – ISM: structure – H II regions – radiolines: ISM. 1 INTRODUCTIONThe ultracompact H II regions (UCHiiR) found deep withinmolecular clouds provide important information on the earlyphases of the interaction of massive stars with their natalenvironment. The disruption of the cloud material by thehypersonic winds and UV radiation fields of these stars is asevere barrier to an understanding of the process of massivestar formation – only by studying the disruption process willanything be learnt about the innermost regions of the pro-tostellar cloud. The disruption process also involves manyimportant problems of gas dynamics. Considerable theoret-ical effort has gone into modelling the varied morphology ofUCHiiR.Most theoretical attention has so far been given to thecometary regions, which comprise about 20 per cent of ob-served UCHiiR (Churchwell 1990). Perhaps the most de-tailed model, that of Van Buren & Mac Low (1992, andreferences therein), treats them as the steady-state partiallyionized structures behind bow shocks driven by the windsof stars moving through molecular cloud material. Althoughit has been argued that some morphologies which are notapparently cometary (in particular, core–halo) can be ex-plained as cometary structures viewed close to their axes(Mac Low et al. 1991), the shell and multiply peaked mor-phologies cannot, and so other models should also be inves-tigated.There are also a number of unresolved questions withregard to the cometary models. First, the star is assumed tobe moving through relatively homogeneous molecular cloudgas. Yet it is well known that cloud material has a clumpydistribution down to very small scales. Secondly, currentcometary models require rather high stellar velocities, char-acteristically 10–20kms