UNRESOLVED WIND-DRIVEN SHELLS AND THE SUPERSONIC VELOCITY DISPERSION IN GIANT H-II REGIONS

The presence of giant shells or loops in giant H II regions is clear w itness to the mechanical energy input from massive stars. Here we eval uate the impact that winds may have on the structure of giant nebulae and on their supersonic velocity dispersion. We follow the recent sugg estion of Chu and Kennicutt to see whether a combination of a large nu mber of unresolved wind-driven shells caused by massive stars could pr oduce the integrated broad Gaussian profiles typical of giant H II reg ions. The results, accounting for a wide range of energies, densities, and velocities or ages of the expanding shells, show that supersonic Gaussian profiles may arise only from a collection of unresolved wind- driven shells if the shells present a peculiar velocity distribution w hich implies a strongly peaked age distribution leading to an awkward star formation history. On the other hand, a uniform distribution of a ges produces profiles with a flat-topped core defined by the terminal shell velocity and a steep decay as upsilon(-6), up to the largest det ectable shell speed. Thus, supersonic profiles can arise only if the f inal speed of the unresolved shells is supersonic. This implies an equ ally supersonic random speed of motions in the ionized gas disrupting the shells before they slow down to subsonic speeds. It also implies a mechanism, independent of the shells caused by massive stars, respons ible for the supersonic stirring of the background medium. These facts , together with the conditions for shells to remain unresolved by pres ent-day devices (energies, final speeds, and ages), indicate that the winds may be produced by low-mass stars. In the latter case, if the so urces move supersonically in the gravitational potential of the whole system, they could stir the gas, with their cometary bow shocks, to a velocity dispersion sigma(gas) similar to sigma(stars), causing a supe rsonic local random speed of motions within the system.