Annu. Rev. Astron. Astrophys. 1992. 30: 575-611 Copyright © 1992 by Annual Reviews. All rights reserved

7. SUMMARY

Impressive advances in both observation and interpretation have been made in the time since the general review of radio emission from spiral galaxies by van der Kruit & Allen (1976). The VLA has provided the speed to map hundreds of galaxies in complete samples, the sensitivity to detect dozens of individual radio supernovae and SNRs in external galaxies, the resolution to map nearby galaxies in detail and to resolve even the most compact nuclear starbursts, and the ability to make maps at several frequencies with similar (u, v)-plane coverage. Equally important is IRAS, whose FIR sky survey provided complete samples of normal galaxies ranging from the quiescent to the ultraluminous (L_FIR > 10¹¹ L) and showed that the FIR/radio correlation is remarkably strong and universal. The FIR/radio correlation is the single most important factor in the widespread acceptance of massive stars as the energy source for both the FIR and nonthermal radio emission from normal galaxies. It also suggests that the star-formation rate SFR(M 5 M) is the dominant variable controlling the range of FIR and radio luminosities observed. Other factors, such as magnetic field strength, either play a minor role or do not vary significantly from ordinary disk galaxies to luminous nuclear starbursts.

Models relating the thermal radio, nonthermal radio, and FIR luminosities of galaxies to SFR(M 5 M) can now be constructed that are consistent with, but not required by, the data. For example, the empirical relation between the supernova rate and nonthermal radio luminosity (Equation 18) does not explain how cosmic rays are generated in supernova remnants, propagate in the interstellar medium, and lose their energy or escape from the galaxy; it merely sweeps these problems under the rug. But such models do highlight the problems that remain to be solved by better analysis and observation. Existing data on the strong nearby source in M82 indicate what might be done with other normal galaxies using improved instruments in the near future. Direct estimates of the supernova rate are badly needed to pin down the relation between nonthermal radio luminosity and the SFR. The VLA is capable of monitoring the number of SNRs in M82, their flux densities, and even their angular sizes. Periodic VLA observations of other nearby galaxies would be sensitive enough to detect new radio supernovae and determine their average radio supernova rate without the optical limitations of confusion by starlight and extinction by dust. At most we have an outline of what might be; only the future will bring the full picture of what must be.

ACKNOWLEDGEMENTS

The National Radio Astronomy Observatory is operated by Associated Universities, Inc., under cooperative agreement with the National Science Foundation. I thank Pat Crane, George Helou, Allan Sandage, Joan Wrobel, and especially Ko Hummel for many valuable suggestions and comments.