7.1. Challenges and Prospects
The two-component model approach is clearly an over-simplification of the complexities in a real system, representing what is almost certainly a continuous distribution over some parameter by a linear combination of the extremes. One challenge ahead is to formulate more realistic descriptions of star-forming galaxies, incorporating a distribution of ISM mass over the key parameters density, intensity and hardness of heating radiation. Additional parameters such as cosmic ray flux, metallicity, dust abundance and properties, or X-ray flux might be added as secondary parameters. From such a description, one would estimate observables as integrals over the phase space defined by the key parameters (see Section 2 above), and translate empirical properties of galaxies into constraints on the ISM properties.
A related challenge is to formulate a realistic description of the geometry of dust, gas and heating stars. The ISM appears clumpy on all scales, perhaps best described as fractal in its structure as evident from IRAS, HI or CO maps of emission (Falgarone & Phillips 1996). Detailed studies of nearby PDRs suggest substantial ultraviolet penetration far from the boundary of the HII region (Howe et al. 1991; see also discussion in Section 2.1 of Hollenbach & Tielens 1997). Geometry may well be a critical factor in some of the observed properties of galaxies; it may well affect the averaging over regions and directions in ways that inject apparent simplicity in otherwise quite complex systems. Detailed models may be the only way to establish the role of geometry in normal galaxies.
Neglecting for now the complexities above, one can expect substantial progress in understanding normal galaxies by exploiting ISO data with new approaches. In addition to the powerful spectroscopic techniques offered by ISO, the improved spatial resolution of ISO-CAM should allow us to go beyond luminosity, and work with surface brightness in the infrared. This is a distance-independent quantity which can be measured and compared in nearby as well as distant galaxies, and combines information on the intensity of heating radiation and the column density of dust, and therefore might help us disentangle these two quantities using additional information from other observables. Some work along these lines based on IRAS data has been published (Wang & Helou 1994; Helou & Wang 1995; Meurer et al. 1997), but ISO data offer much better sensitivity and spatial resolution (Dale et al. 1999; Dale et al. 2000a).
With ISO data in hand, it should be possible to answer some basic questions about galaxies, deriving for instance an accurate star formation rate from a combination of observables, and attaching specific physical meaning to standard paramaters commonly used today, such as IRAS colors, line-to-continuum ratios, or infrared-to-blue ratios. More difficult but probably within reach of ISO data interpretation would be a means to distinguish between physical parameters and mixing ratios, so one can tell whether two galaxies differ because their PDRs are systematically different, or because they have a different mix of PDR, HII regions and molecular clouds.