Astron. Astrophys. 327, 1185-1193 (1997)

7. The detectability of globules in HII regions

In an idealized situation, the perfect balance established between the number of recombinations and the stellar uv output that causes an HII region implies that high density condensations, or globules, able to trap the ionization front are not always detectable if placed at particular locations with respect to an observer.

We refer to Fig. 11, where a globule is seen in three different locations within an HII region. In case a, we look into an HII region with radius R, excited by a central OB star. We consider two adjacent lines-of-sight, where line 1 passes through the entire HII region, and line 2 is intercepted by a neutral globule located behind the OB star at a distance r from the star. The uv radiation penetrates a distance s into the globule. We assume that the ambient gas and the globule are homogeneous, with densities of and . Along line 1 the number of recombinations is proportional to . Along line 2 the number of recombinations is proportional to . In the particular geometry where the globule is right behind the star, the radiation available at the globule surface equals the recombinations that would have occurred behind the globule , that is equals . The emission measure along line 1 is , and along line 2 it is

Fig. 11. A schematic representation of a globule seen in three different locations within an HII region. The globule is black in each panel, and its bright rim is white, with a depth s and a width l. The observer is towards the bottom, and four different lines-of-sight are considered

in other words the emission measures of the two adjacent lines-of-sight are identical, and the globule is invisible! From the above equality, and defining the density contrast , it follows that

that is, for globules closer to the star the penetration into the globule will be deeper. At the same time, denser globules will allow a uv penetration distance inversely proportional to the square of the density contrast. Scattering by dust particles around the borders of a globule will enhance its possible detection although only a small contrast is expected from lines of sight traversing the whole HII region.

The chances of detection increase if the ionized jacket around the globule is seen tangentially rather than face on. In this case, globules would appear as sharp luminous edges or bright rims, against the emission of the HII region (see Fig. 11, line 3). Here the emission measure is increased over adjacent lines-of-sight by .

Finally, if globules are placed between the observer and the ionizing source (Fig  11, line 4), they will be apparent as dark patches against the luminous background, because they block the HII region emission , as well as the bright rim contribution .

In a more realistic situation, detection strongly depends on the shape of the high density condensations, which could go from small round globules to elongated elephant trunks. It depends also on the number and location of the massive stars causing the ionization. The first issue defines the shape of the observed bright rims around condensations as well as the shape of the detected dark patches seen against the luminous background. The number of stars and their location strongly define the general orientation of bright rims in an HII region. Detection of bright rims is also enhanced by the gas dynamical evolution of the ionized globule material which enhances the surrounding gas density as it streams away from the condensation in a localized champagne flow (Bedijn & Tenorio-Tagle 1984). This flow leads to a fuzzy but broader rim around high density condensations, perhaps better seen around condensations sitting between the observer and the ionizing sources. This is very noticeable in the case of Thackeray 1. The broad rim effect is enhanced if the streaming of matter away from the ionization front is forced to converge. As the flow of ionized gas follows pressure gradients, and these are inevitably perpendicular to the border around dense condensations, the flow resultant from an ionized concave dense edge would lead to convergency of the ionized flow. The opposite is to be expected from the ionization of convex cloud surfaces, since then the pressure gradient would cause the flow to diverge, making the broad bright rim effect less evident. The enhanced rim luminosity due to a converging flow is quite clear in the concave borders of Thackeray 1 and 9. On the other hand, round and convex edges leading to a diverging flow present a much less brightened edge, as in some edge sectors of Thackeray 1 and 9 as well as in 5, 6 and 7 etc.

© European Southern Observatory (ESO) 1997

Online publication: April 6, 1998
helpdesk.link@springer.de