The topics covered in this article are both very old and very new. It has been known for over three decades that a large segment of the galaxy population exhibits signs of unusual activity in their nuclei, and for nearly as long people have puzzled over the physical origin of this activity. Over time the observational material at optical wavelengths has improved markedly, especially with the completion of the Palomar survey, but the debate has only intensified. Given their abundance, most of the controversy has centered, not surprisingly, on LINERs. While the nonstellar nature of a sizable fraction of LINERs is now incontrovertible (e.g., those with broad H emission), the AGN content in the majority of the class remains unsettled. Determining the physical origin of these systems is more than of mere phenomenological interest. Because LINERs are so numerous - being the dominant constituent of the local LLAGN population and a sizable fraction of all galaxies - they have repercussions on virtually every issue related to AGN and BH demographics.
The main source of contention stems from the fact that Mother Nature knows too many ways of generating nebular conditions that qualitatively look similar to the low-ionization characteristics of LINERs at optical wavelengths. A dizzying array of excitation mechanisms has been proposed to explain LINERs, ranging from variants of conventional AGN photoionization, to shocks of various flavors, to interstellar processes such as cooling flows and turbulent mixing layers, to stellar-based photoionization by populations both young and old, prosaic and exotic. This field has suffered not from a shortage of ideas, but from too many. As a consequence, whenever LINERs are discussed, it is customary to end with the pessimistic mantra that they are a mixed-bag, heterogeneous class of objects, a statement with somewhat dismissive connotations that is often taken to mean that we have no idea what they are and that they are too messy to deal with. This is an unfair characterization of the progress that has been made, and I think that there is good reason to sound a more positive note.
As summarized in this review, a number of developments during the last few years shed considerable light on the physical origin of LLAGNs in general and LINERs in particular. The key advances have come from the broader perspective afforded by observations outside of the traditional optical window, especially in the radio and X-rays, although important insights can also be credited to optical and UV data taken with HST. In all instances, high angular resolution has been a critical factor to disentangle the weak nuclear emission from the blinding host galaxy background.
Other developments have been instrumental in forging a coherent view of nuclear activity in the nearby Universe. On the theoretical side, rapid advances in the study of radiatively inefficient accretion flows, originally primarily motivated by applications to X-ray binaries and to the Galactic Center source Sgr A*, has led to a growing appreciation that they are also relevant to LLAGNs in general. Many investigators have sharpened the physical analogy between the spectral states of X-ray binaries and certain classes of AGNs, an effort that has resulted in a more holistic picture of BH accretion, especially as it concerns the evolution of the accretion flow in response to variations in mass accretion rate and the mechanism for generating jets or outflows. Meanwhile, the dynamical detection of supermassive BHs, their ubiquity, and the discovery of scaling relations between BHs and their host galaxies have given a major boom to studies of AGNs in all their multi-faceted manifestations. More than ever, in the grand scheme of things, AGNs are no longer viewed as rare and exotic oddities but as natural episodes during the life cycle of galaxies during which their BHs accrete, grow, and shine. The impact of BH growth and AGN feedback have emerged forcefully as major new themes in galaxy formation. LLAGNs gain an even greater prominence within this context. Although the bulk of the mass density of BHs was accreted in a luminous, radiatively efficient mode, it behooves us to understand how BHs spend most of their lives. The detection of supermassive BHs has also fundamentally altered the character of the discourse on LLAGNs. We can now shift our attention from the question of whether LLAGNs contain BHs - an implicit or explicit motivation for much of the past discussion on the nature of these sources - to why these BHs have the properties that they do. Among other things, LLAGNs can be used as an effective platform for exploring accretion physics in highly sub-Eddington systems and for investigating physical processes in the circumnuclear regions of galaxies that are normally masked by brighter nuclei.
The following is a list of the "top ten" results from this paper.
Approximately 2/3 of local E-Sb galaxies exhibit weak nuclear activity incompatible with normal stellar processes; in contrast, only about 15% of Sc-Sm galaxies show AGN activity (Section 3).
The vast majority of LINERs, and, by implication, most nearby weakly active nuclei, are genuine, accretion-powered AGNs (Sections 6.1, 6.5).
The ubiquity of LLAGNs in galaxies with bulges strongly supports the current paradigm derived from dynamical studies that all bulges contain BHs. However, the detection of AGNs in some bulgeless, even dwarf, galaxies proves that bulges are not necessary for the formation of central BHs (Section 7).
The luminosity function of nearby LLAGNs follows L-1.2 ± 0.2 from LH 3 × 1041 to 1038 ergs s-1, below which it appears to flatten down to LH 6 × 1036 ergs s-1 or MB -8 mag (Section 5.9).
Stellar photoionization by young or intermediate-stars and shock heating can be ruled out as the excitation mechanisms for LLAGNs (Sections 6.2, 6.3).
Despite the overall success of AGN photoionization models, many LLAGNs, especially type 2 sources, have a shortage of ionizing photons. The energy deficit problem could be solved with cosmic ray heating and extra ionization from evolved (post-AGB) stars, diffuse thermal plasma, and the cumulative X-ray emission from low-mass X-ray binaries (Section 6.4).
Variations in the mass accretion rate give rise to the different classes of emission-line nuclei. LINERs are the low-luminosity, low-accretion rate extension of Seyferts, followed by transition nuclei, and ending with absorption-line nuclei at the end of the BH starvation sequence (Section 6.5).
LLAGNs are not simply scaled-down versions of powerful AGNs. Their central engines undergo fundamental changes when the accretion rate drops to extremely sub-Eddington values. In this regime, the BLR and obscuring torus disappear (Sections 5.5, 5.6). LLAGNs do not follow the standard AGN unification model.
Below a characteristic luminosity of ~ 1% Eddington, the canonical optically thick, geometrically thin accretion disk transforms into a three-component structure consisting of an inner vertically thick and radiatively inefficient accretion flow, a truncated outer thin disk, and a jet or outflow (Sections 5.8, 8.3).
At the lowest accretion rates, an increasing fraction of the accretion energy gets channeled into a relativistic jet. The emitted energy is mainly kinetic rather than radiative. Since radiation and kinetic jets interact differently with the surrounding gas, this has important implications for AGN feedback into galaxy formation (Section 8.2).
My research is supported by the Carnegie Institution of Washington and by NASA grants from the Space Telescope Science Institute (operated by AURA, Inc., under NASA contract NAS5-26555). I would like to recognize my collaborators who have contributed to the work covered in this review, especially A.J. Barth, M. Eracleous, M.E. Filho, A.V. Filippenko, J.E. Greene, D. Maoz, E.C. Moran, C.Y. Peng, A. Ptak, E. Quataert, H.-W. Rix, W.L.W. Sargent, M. Sarzi, J.C. Shields, Y. Terashima, J.S. Ulvestad, and J.M. Wrobel. Several of them (A.J. Barth, M. Eracleous, J.E. Greene, M. Sarzi, J.C. Shields, Y. Terashima) read an early draft of the manuscript and provided useful feedback that helped to improve it. Some of the concepts expressed in the review were sharpened after correspondence with M. Eracleous, G. Ferland, J.C. Shields, and Y. Terashima. I thank A.J. Barth, A.V. Filippenko, and W.L.W. Sargent for permission to cite material in advance of publication, H.M.L.G. Flohic and M. Eracleous for providing the images for Figure 5, and Salvador Ho for drafting Figure 13. I am grateful to J. Kormendy for his steadfast encouragement, wise counsel, and meticulous editing.