Because the peak of the thermal dust emission occurs in the far-infrared (60-200 µm) for Galactic interstellar clouds, this wavelength domain is recognized as a very important `window` to study the interstellar medium, both in the Milky Way and in external galaxies. However, due to the lack of spatial resolution provided by telescopes operating at these wavelengths, the interpretation of the signals measured in external galaxies is still a subject of controversy. The debate is focussed on the origins of the heating of dust grains and of the C+ line at 158 µm. This fine structure line of ionized carbon is one of the main cooling lines of neutral atomic gas (Wolfire et al. 1995), but it is also seen in HII regions, with different excitation mechanisms in these two cases. Other sources of C+ emission are the dense photodissociation regions (PDRs) where intense UV radiation from massive stars impinges on molecular clouds. In these regions, the photodissociation of carbon monoxide creates a layer of ionized carbon at the edge of the molecular cloud, which is a source of intense C+ emission.
The far infrared emission is produced by dust grains heated by star light but the relative contribution of young massive stars on one hand, and of the bulk of the stellar population on the other hand, are still debated (see e.g. Walterbos & Greenawalt 1996, Devereux & Young 1990, Cox & Mezger 1989, Thronson et al. 1990, Rice et al. 1990). Using low resolution data from IRAS, IR colors have been used to assess the respective roles of star forming regions and cirrus in external galaxies at a global scale (e.g. Calzetti et al. 1995). It is difficult to find spiral galaxies dominated by a single population of heating sources, since the relative contributions of massive stars and of the disk population vary from galaxy to galaxy, and probably from place to place in a given galaxy. A main problem of these studies is the lack of spatial resolution at far infrared wavelengths. Even when it is possible to use high spatial resolution data at millimeter and submillimeter wavelengths, the gap in spatial resolution from the local interstellar clouds to external galaxies is huge: for nearby galaxies located at distances D of a few Mpc, a one arcmin beam encompasses , much larger than the size of a single molecular complex in the Galaxy, whereas structures down to 0.01 pc are commonly observed in local clouds (Falgarone et al. 1992) and in the Magellanic Clouds (Rubio et al. 1993).
Numerical simulations provide a way out of this problem: it is possible to reproduce the observed molecular gas distribution of nearby spiral galaxies with numerical simulations using observed data as input parameters. For example Garcia-Burillo et al. (1993) were able to fit the spatial distribution and kinematics of the molecular gas in M 51 using the cloud collision code developped by Combes & Gerin (1985). Though this code has a spatial resolution of a few hundred parsecs (the cell size for the large scale dynamics), it is possible to include "micro-physics" at the parsec scale inside each cell. We have taken this approach and implemented star formation in this code to study the far infrared and C+ emissions in the spiral galaxy NGC 946. The next section summarizes the current data on NGC 6946. We describe the model in Sect. 3 and present the results for NGC 6946 in Sect. 4. The implications of this work for the interpretation of the C+ and FIR emission of spiral galaxies are discussed in Sect. 5.
© European Southern Observatory (ESO) 1998
Online publication: September 30, 1998