Next Contents Previous

4.3.1. The Global Infrared Spectrum

Turning now to the long wavelength end of the spectrum, photometry at 120-200 µm using ISO-PHOT is starting to constrain the distribution of dust temperatures at low heating levels, especially in nearby well resolved galaxies such as M31 (Haas et al. 1998, 1999), M51 or M 101 (Hippelein et al. 1996), where cold dust dominates the luminosity. Similar analysis on more active galaxies is also under way (e.g. Klaas et al. 1997; Klaas, Haas & Schulz 1999) to obtain the best possible estimates of the total infrared emission and therefore of the dust mass. Alton et al. (1998) have reported that the emission is more extended at 200 µm than at shorter wavelengths in several galaxies. This result however hinges closely on a very precise knowledge of the beam shapes at various wavelengths, which was not yet achieved at the time of that publication.

Figure 7

Figure 7. The synthesized model spectra of Dale et al. (2000a), showing the progression of shape from quiescent to active star-forming galaxies. The same set is plotted here as summarized in Table 1. The spectra are plotted so the Aromatic Features are scaled to the same amplitude just for ease of graph-reading, these features being the most stable part of the spectra. Note the march of the peak towards shorter wavelengths as activity increases. These synthesized spectra do not account for the general correlation between R(60, 100) and the infrared-to-visible light ratio.

A proper determination of the total infrared luminosity and the long-wavelength spectral shape in normal galaxies is critical to estimating the contribution of galaxies to the infrared and submm extragalactic background light, and thereby deriving the infrared term in the star formation history of the Universe. While work on the ISO-PHOT calibration continues, one could estimate an improved infrared spectral energy distribution by combining the mid-infrared results described above with existing IRAS data (Figure 7). Such a rough estimation is presented in Table 1 for several different levels of activity in galaxies, parametrized by the 60-to-100 µm ratio. It should be noted that all information at lambda > 100 µm in Table 1 is based on modeling IRAS data using a power-law distribution of dust mass as a function of heating intensity, and does not use any empirical constraint (Dale et al. 2000a; Helou et al. 2000). Table 1 illustrates the general trend, but also ignores variations in spectral shape at constant 60-to-100 µm ratio, including intrinsic scatter in the ratio of mid-infrared to far-infrared (Lu et al. 1999), and dispersion in the 25-to-60 µm ratio. The 20 to 40 µm range appears to show the most significant growth in fractional terms at the expense of the submillimeter as the activity level increases, suggesting that the 20-40 µm continuum may be the best dust emission tracer of current star formation in galaxies. Even after ISO, our knowledge of the detailed shape of the spectrum at lambda > 100 µm remains model-dependent, and may not improve significantly until the launch of SIRTF, and then of FIRST (Far infraRed and Submillimeter Telescope).

Table 1. Rough energy distribution across the infrared spectrum for galaxies with various levels of star formation activity, parametrized by the IRAS color in the first column. Each column heading gives the wavelength range over which the spectrum is integrated, and the table entries are the fractions of total infrared luminosity appearing in that range. The spectral range in column 6 corresponds to the ``FIR'' synthetic band (Helou et al. 1988).

fnu(60 µm)/fnu(100 µm) F(3 - 5 µm) F(5 - 13 µm) F(13 - 20 µm) F(20 - 42 µm) F(42 - 122 µm) F(122 - 1100 µm)

0.40 0.024 0.122 0.033 0.10 0.41 0.32
0.63 0.017 0.086 0.037 0.18 0.50 0.19
1.00 0.008 0.048 0.043 0.28 0.54 0.09

Next Contents Previous