Annu. Rev. Astron. Astrophys. 2000. 38: 761-814 Copyright © 2000 by Annual Reviews. All rights reserved

5. CONCLUSIONS AND OUTLOOK

ISO has provided progress on many fronts. The impulsive heating mechanism of single UV (and optical?) photons impinging on very small dust grains/large molecules and emitting in the near-IR/mid-IR now appears firmly established. The mid-IR SED is a valuable diagnostic of the activity level in galaxies. The mid-IR emission from early-type galaxies is composed of stellar, interstellar, and AGN contributions which can now be separated. In nearby spirals the far-IR data require cool equilibrium dust with a dust-to-gas ratio similar to that in the Milky Way. ISO has added tantalizing results on the nature of the halos and outer disks of spiral galaxies, but has not solved the overall puzzle of the dark matter residing there.

Starburst Galaxies (and the Starburst/AGN Connection):

ISO has added infrared spectroscopy and spectrophotometry to the astronomer's arsenal of powerful analytical tools. Dust-enshrouded AGNs and starburst galaxies can be qualitatively and quantitatively distinguished. ISO has confirmed that infrared luminous starburst galaxies typically have low nebular excitation. The likely explanation is that starbursts are episodic and rapidly aging. Star formation provides strong negative feedback that tends to quench active starbursts in little more than an O-star's lifetime.

AGNs:

The ISO data generally support unified accretion disk+circum- nuclear torus/disk models. Emission from AGN-heated, thick tori or warped disks dominates the mid-IR emission of QSOs, radio galaxies, and Seyfert nuclei. Based on a combination of mid-IR spectrophotometry, radio/IR imaging and radiative transport models, star formation likely plays an important role in accounting for the far-IR emission of many Seyfert galaxies. However, the issue of what powers the (thermal) far-IR emission of QSOs and radio galaxies remains uncertain. Mid- and far-IR spectroscopy of powerful AGNs with SIRTF and SOFIA will provide the next major step in this field. A promising new method has emerged for reconstructing the (hidden) EUV SEDs of AGNs from optical/infrared coronal line ratios. SIRTF will tell how useful this tool is for quantitative studies of different classes of AGNs.

(U)LIRGs:

With ISO, much progress has been made in understanding the properties, energy sources, and evolution of (ultra) luminous infrared galaxies, which are a spectacular curiosity locally but very common and important in the early Universe. (U)LIRGs are composite objects containing both powerful AGNs and starbursts. Star formation plausibly dominates the luminosity of most sources. At the highest luminosities, AGNs take over. The standard paradigm of ULIRGs as a (late) phase in the merger of two gas-rich galaxies is confirmed, but their evolution is more complex than was considered before. ULIRGs do not seem to undergo an obvious metamorphosis from a starburst-powered system of colliding galaxies to a buried and then finally naked QSO at the end of the merger phase. Local effects play an important role. SIRTF will increase the statistics and quality of the mid-IR spectroscopy, and will also be able to apply the new IR spectroscopic tools to higher redshift sources. XMM - Newton and Chandra will do the same at hard X-rays. SOFIA promises sensitive far-IR spectra to help tackle the problem of the weak far-IR line emission seen by ISO in nearby ULIRGs.

High-z, Dusty Star Formation:

COBE, ISO, and SCUBA have clearly demonstrated that dusty star formation plays an important role in the early Universe. This is a very exciting development. However, the present data have barely scratched the surface of this important emerging field, and their interpretation is still in a stage of infancy. New models and follow-up work are in progress, and updates give an increasingly detailed picture. ISOCAM has identified an important population of active starburst galaxies (mainly LIRGs) at z < 1.5. This population accounts for 10-60% of the far-IR/submm background. ISOPHOT has begun to detect a ULIRG component at moderate redshift (z ~ 0.2 to > 1) that may account for an additional 10% of the background. Submm observations (SCUBA, etc) are detecting a component of high redshift (z 2) ULIRGs, which make up the remainder of the background. Deeper surveys with SIRTF, SOFIA, FIRST, and ALMA will be one prime objective of the next decade. Identifying counterparts at other wavelengths and studying their nature will be another. The surveys will resolve most of the far-IR/submm background into individual sources. IR spectroscopic follow-up (with ground-based telescopes, SIRTF, FIRST) and hard X-ray observations (Chandra, XMM - Newton) will establish the redshifts and will determine what fraction of the IR sources comprises active starbursts and what fraction comprises dusty Seyferts/QSOs. We can look forward to an exciting time!

ACKNOWLEDGMENTS

We would like to thank the many colleagues who have sent us reprints or preprints, or who have given us input, comments, and advice: in particular, P Andreani, H Aussel, R Barvainis, J Clavel, P Cox, H Dole, D Elbaz, J Fischer, A Franceschini, M Haas, G Helou, D Hollenbach, M Kaufman, O Laurent, D Lemke, S Lord, M Luhman, D Lutz, R Maiolino, S Malhotra, H Matsuhara, F Mirabel, AFM Moorwood, T Nakagawa, H Okuda, J-L Puget, D Rigopoulou, B Schulz, S Serjeant, B Smith, M Stickel, E Sturm, Y Taniguchi, M Thornley, and D Tran. We would like to thank Susanne Harai, Dieter Lutz, Linda Tacconi, and Lowell Tacconi-Garman for help with the manuscript and for useful comments. We are especially grateful to F Mirabel and AFM Moorwood for their comments on the manuscript.