3.1.4. Other Estimators
Of many other parameters used in analyzing the empirical properties of galaxies from IRAS data, only two are shown here, for illustration rather than completeness. In using such estimators, one should be aware of the assumptions that enter into relating them to specific physical quantities, and estimate the uncertainties resulting from the inaccuracy of the assumptions in addition to the inherent measurement uncertainty.
The ratio of f(100 µm) to the HI 21cm line flux has been used as an indicator of the dust-to-gas ratio. This interpretation however works only in very broad terms when comparing galaxies, since several factors other than dust to gas ratio affect the parameter (Helou 1985). For instance, the 100 µm-to-HI ratio scales at least linearly with heating intensity, and depends on the assumed ratio of atomic to molecular hydrogen, as well as on the fraction of HI-related dust which is heated sufficiently to emit at 100 µm, as opposed to dust in an extended HI envelope which remains cold enough as to be irrelevant at 100 µm. This estimator works better locally within galaxy disks as a ratio of surface brightness at 100 µm and in HI as opposed to a global parameter (Dale et al. 1999). See Melisse & Israel (1994) for an interesting analysis.
The ratio of FIR flux to CO flux has been used as a measure of star formation efficiency, in the context of interpreting L(FIR) as a measure of star formation rate, and L(CO) as a measure of the gas mass available for star formation. The inverse quantity can also be interpreted as an indicator of how long the current star formation rate can be sustained by the system. The caution here is that this estimator is useful only to the extent allowed by the uncertainties in interpreting L(FIR) and L(CO) as just stated. L(CO) in particular traces CO rather than total H2 mass, and includes a significant dependence on the CO excitation conditions, as well as optical depth effects (Maloney & Black 1988, Aalto et al. 1994).
It is clear that much caution needs to be exercised in using all these estimators. Using several simultaneously helps in narrowing down the uncertainties of interpretation. For example, Knapp et al. (1987) combined f(100 µm) / F(HI) with f(100 µm) / F(CO) to trace the origin of the infrared emission. The discussion above illustrates the importance to the study of normal galaxies of understanding ISM physics and star formation processes on scales from a parsec to a kpc. This understanding must be derived from empirical and physical modeling studies of the Milky Way and Local Group galaxies, at the best available spatial and spectral resolution.