Active galactic nuclei

By: David Bradley

The following article by David Bradley

is a summary report of a meeting on the subject of active galaxies and quasi-stellar objects held recently at the Royal Society. For information on RS meetings please visit their website - http://www.royalsoc.ac.uk

Galactic impact and universal truths

Astronomers recognized almost fifty years ago that stars have an enormous impact on the space around them, affecting the space in between and synthesizing the heavy elements from which planets are made. Active Galactic Nuclei (AGN) are known to release vast amounts of energy sometimes outshining all the other stars in a galaxy combined but their role in affecting their surroundings is only now being discovered.

AGN release vast amounts of energy, but unlike their non-active counterparts a large proportion of this energy is not emitted by the familiar components of galaxies, the stars, dust, and interstellar gas. Instead, the energy emerges across the electromagnetic spectrum from the infrared to the gamma ray and comes from the core of the galaxy. Such galaxies are usually referred to as AGN because the majority appear to be powered by their central region. AGN often spit out jets of matter that stretch thousands of light years into space and provide the power supply for other astronomical species such as radio galaxies and radio-loud quasars.

The effect of AGN on the overall structure of the Universe is not dramatic, explained Simon White of the Max Planck Institute for Astrophysics in Garching, Germany, in his opening remarks, however, on a smaller scale research has begun to reveal that AGN are very much active in structuring the universe in ways that were not fully appreciated before.

It comes as a surprise to many people to learn that astronomers believe that almost all massive galaxies have at their centre a massive black hole. Indeed, astronomers have used spectral studies to reveal that almost half of our neighbouring massive galaxies have an active core; they are AGN.

Moreover, the relationship between the galaxy and the black hole at its core has been found to be much closer than earlier observations suggested. The mass of the black hole is tightly related to the overall mass of the galaxy. The initial formation of the black hole and the galaxy itself are also known to be inextricably linked. Astronomers would like to explore the nature of this link and find answers to the question: Is the mass of the black hole controlled by the galaxy or is it the other way round? The answer would help astronomers understand whether it is a black hole that leads to the accumulation of matter that produces the galaxy, or whether the matter in a galaxy ultimately leads to the formation of the central black hole. Alternatively, the galaxy and the black hole may form concomitantly one feeding off the other.



The effects of active galaxies on the universe also raises many questions. The ionising radiation from an AGN can affect the structure of the space between galaxies, the inter-galactic medium, by heating it and changing its ionisation. Were black holes responsible for the early universe lighting up after the period known as the Dark Ages just 470 million years after the Big Bang? Or, was it star formation itself? Astronomers do not yet know.

Data from rapidly advancing astronomical techniques that span the electromagnetic spectrum covering radio waves, infrared radiation and X-rays are being combined with increasingly realistic simulations of structure and galaxy formation allowing researchers to address these questions. The answers they are finding will take our understanding of galaxies, star formation, and ultimately the early universe forward by light years from where it stands now.

Circumstantial black holes

Scott Tremaine of Princeton University began with the now-accepted assumption that AGN are powered by massive black holes. The evidence for this assumption, he explained, lies in such properties as the persistence of the enormous jets of matter from AGN and the fact that almost all dense astronomical systems ultimately evolve into a massive black hole anyway.

Detection of these exotic objects, or proof that they don't exist, will provide insight into the nature of AGN, the formation of galaxies, the properties of the central regions of galaxies, and even help physicists demonstrate that Einstein's theory of general relativity is right under the extreme conditions found near black holes.

One prediction that has been borne out by recent data from the Hubble Space Telescope and ground-based telescopes is that we should find extinct AGN closer even than the well-known Virgo cluster. It certainly now appears that most galaxies indeed do have a massive black hole at their centre. The classic and best- known case is our own galaxy the Milky Way. As such, we have the very object astronomers want to study in distant quasars essentially on our doorstep.

Astronomers have used techniques such as stellar kinematics, gas kinematics, and maser kinematics to look for evidence of black holes in nearby galaxies. These techniques essentially probe the velocities of stars close to the centre of a galaxy. The measurements reveal that there are many cases of stars orbiting a central mass so rapidly that it must be supermassive, that is a million to a billion times the mass of the sun, yet very small; it has to be a black hole.

The evidence for black holes at the centre of such galaxies is still accruing. Tremaine emphasized that there are about 30 believable measurements of black hole masses but he stressed that the evidence points only to a massive, dense dark object at the centre; we cannot prove that this is a black hole but we know of nothing else in the universe that could have such properties. Irrespective of that tiny doubt, AGN require a black hole to behave the way they do. The only plausible candidate is a black hole.

There are many key questions that remain unanswered asserted Tremaine, such as what determines the growth of black holes and what happens to the black holes at the centres of merging galaxies? Moreover, why is black hole mass so tightly correlated with the galaxy's properties?

Galactic demographic

If we want to be able to understand the impact of AGN on the formation and evolution of galaxies then it would help if we could study the populations of AGN and the population of galaxies together, proclaimed Guinevere Kauffmann of the Max Planck Institute for Astrophysics at Garching. That, she said, has recently become possible on a large scale thanks to advent of large sky surveys, such as the Sloan Digital Sky Survey (SDSS).

Kauffmann, who collaborates with Johns Hopkins University researcher Tim Heckman, explained how such surveys can help astronomers see how AGN fit into the galaxy population as a whole. The survey contains more than 100,000 galaxies, which has been used to study how galaxies are distributed over very large scales, and among these can be found many AGN.

The study of galaxies in the local universe, explained Kauffmann, began with twentieth century astronomer Edwin Hubble who was the first to study how the colours and shapes of galaxies are related. However, the large surveys allow astronomers to look closely at the real statistical distribution of these properties, and by plotting any one property against any other the most striking characteristic that emerges is that the galaxy population falls into two basic families; the population is bimodal, with blue galaxies and red galaxies.

Red galaxies tend to be found in groups and large clusters, while blue galaxies are found less tightly packed in space. This family split is also seen in terms of size of galaxies, galactic shape, in the number of stars they contain and in the age of the stars in the galaxy. Massive galaxies containing many stars have high densities and older stars, for instance. These points arise repeatedly when the SDSS is used to plot characteristics of galaxies.

This bimodality extends to whether galaxies are active or passive. Astronomers, such as Kauffmann, have used sophisticated statistical analyses to look closely at the family split among galaxies. She emphasised that AGN have so far not been found in low mass galaxies. If they were there, we would have detected them, she said. The key to detecting AGN, of type 2, which absorb radiation that emitted by the accretion disk that surrounds the black hole, lies in the accurate decomposition of the spectrum for each galaxy in the survey into its absorption and emission lines. This allows the astronomers to distinguish between AGN and star-forming galaxies.

The demographics of the sky survey reveals that AGN reside almost exclusively in galaxies with masses greater than 10 billion suns, and also structural properties like those of "early-type" galaxies. Early type galaxies include elliptical and lenticular galaxies and are round or elliptical in outline, they have often been found to be the hosts of quasars and powerful radio sources and so provide a useful starting point when looking for AGN.

Starburst!

Black hole formation is understood to be a fundamental part of the formation and early evolution of elliptical galaxies and the bulges of spiral galaxies, said Tim Heckman, of Johns Hopkins University, Baltimore, who collaborates with Guinevere Kauffmann of the MPI for Astrophysics. However, he would like to better understand the underlying astrophysics, and asked whether we might use the local universe as a laboratory to explore the black holes we see today. This could provide us with a clearer picture of the processes that occurred in galaxies that are much further away in the early universe.

Astronomers can now detect galactic systems containing AGN and young stars just ten millions years old using indirect methods such as spectroscopy. The Hubble Space Telescope has on-board the necessary instrumentation for making observations that have allowed astronomers to see a scaled-down version of intense star formation - starbursts - in local systems with many young stars and an AGN. Laying bare the starburst-AGN connection is helping us explain the early evolution of galaxies more clearly.

The most massive black holes at the centre of the most massive galaxies are usually dead. They are no longer growing and exist in galaxies without much star formation. Nevertheless, strong AGN require a large black hole mass and an abundant fuel supply in the form of gas to power their activity, emphasised Heckman.

He and his colleagues have carried out statistical analyses of Sloan Digital Sky Survey (SDSS) data, which he explained suggests that such strong AGN are found in relatively massive but "young" galaxies. For these systems, black hole growth time and galaxy building time track each other with respect to black hole mass. The data also reveal that the ratio of black-hole growth and star formation in such massive galaxies is similar to that observed in the past. This, Heckman said, suggests that the evolution of these local galaxies mirrors the much more distant, and hence, much earlier galaxies observed in the early universe.

The core of the problem

Katherine Blundell of Oxford University asked the question: why is it important to account for the enormous jets of matter that are spat out from the central nuclei of some galaxies? The answer to this question will help us understand how the evolution of radio galaxies, the very sources of those jets, affects their surroundings by exuding energy in this way. Some astronomers have even suggested that the merging of two galaxies could be the trigger for the formation of radio galaxies and ultimately jet activity.

It is, she said, important to understand the demography, the number and distribution of radio galaxies, if we are to piece together the role they play in structure formation in the universe. Moreover, there is reason to believe that there might be many examples "out there" that cannot see with current technology.

We already have a robust physical model for the strongest radio sources. Coincident with the active nucleus itself is a compact radio core and emanating from this core are enormous jets of plasma, fully ionised matter, moving at close to light speed. When these jets hit the intergalactic medium they form regions of enhanced surface brightness; so-called hot-spots. Ultimately, these hot-spots influence the way in which energy spreads out through the plasma.

Blundell emphasised that this model could be used to invert backwards to reveal how radio sources are born. It is important to get the details right, she added, as the powerful radio galaxies provide our best shot at understanding how luminosity changes and so will reveal information about galactic evolution. Such research that focuses on the major role of the hot-spots will also help solve the enigma of our lack of observations of large, powerful radio sources first suggested more than two decades ago.

The problem seems to arise as the luminosity of the radio sources falls rapidly for older galaxies and so they are not detected by the sky surveys because the old galaxies are simply not bright enough. The recent recognition of this missing population of older galaxies, previously considered "radio quiet", could help us reassess the data obtained so far. Blundell suggested that a new low-frequency survey with the giant metre-wave radio telescope (GMRT) in India, would yield exciting new results.

Fuelling galactic machines

Despite a faltering start over terminology and acceptance, most astrophysicists now generally agree that AGN, including quasars, are powered by the assimilation, or accretion of gas in to a massive black hole, explained Roger Blandford of Stanford University. Indeed, it now appears that every major galaxy hosts a massive black hole, between a million and a billion times the mass of the sun, in its nucleus.

In the modern parlance, black holes are described as machines, said Blandford. As such, astrophysicists often discuss them in the terminology of engineering referring to energy, input and output, efficiency, and so on. A black hole is a “machine”, explained Blandford, in the sense that it takes in gaseous fuel and converts it to heat, radiation and work, the consequences of which can be observed directly with astronomical telescopes.

Blandford further explained that most of the growth of a black hole occurs when the rate of accretion is close to the so-called Eddington rate. The Eddington limit is the point at which radiation from an object is counteracted by its gravity. At this rate, matter is being assimilated by the black hole from its surroundings at a rate ranging from one percent of a sun to ten suns per year, depending upon the mass of the black hole. When gas is supplied at a rate much above or below the Eddington rate, most of it is blown away in the form of a gaseous wind. Overall the power associated with this wind, the radiation and relativistic jets is comparable. The radiation and wind might be released in any direction while the jet is strongly beamed.

Such enormous energy transfers cannot fail to affect the systems in which they occur and the evolution of AGN should match the growth of holes. The environmental impact of AGN is thus a major factor in galaxy formation and evolution. New observational approaches, will look for additional evidence of this.

Galactic outpourings

Accreting black holes produce powerful outflows, stated Mitchell Begelman of the Joint Institute for Laboratory Astrophysics at the University of Colorado. Sufficient observational and theoretical evidence has accrued that most astrophysicists hold this true. Observations of jets from radio galaxies and spectroscopic evidence in the form of broad absorption lines (BAL), suggest the existence of these outflows. Theory says that the accretion process around a black hole leads to heating and outflow, unless radiative losses are very efficient. The existence of magnetized coronae and winds would further imply an outflow of energy from a black hole.

Jets and winds are significant channels for the energy loss from growing black holes. Current evidence points to the fact that these outflows carry more energy than had previously been thought. These vast outflows of energy cause mechanical heating of their surroundings as shockwaves blast into the surrounding space, as well as gentler forms of heating. The nature and distribution of this mechanical heating by central AGN in clusters of galaxies, however, is not yet fully understood.

Begelman described how numerical computer simulations and semi-analytic models are being used to recreate the observed conditions and help explain the heating processes, which might then account for some of the observed behaviour of AGN. Indeed, energy released might provide enough feedback to break apart the matter that is forming a galaxy at the same time a black hole is growing. This effect might therefore regulate the formation of the galaxy and the growth of the black hole and help explain the observed correlation between the two.

One relatively gentle heating mechanism that has been proposed is so-called "effervescent heating". This might counterbalance the radiative losses in the central regions of clusters. It may also heat up the surrounding gases more than can be accounted for by the heating effects of gravitation as smaller clusters merge to form larger ones. Begelman explained how effervescent heating might therefore account for the disorder, or excess entropy, observed even at great distances, up to several million light years, away from the AGN itself.

Hot oscillations

Jeremiah Ostriker of the Universities of Cambridge and Princeton had one message for the meeting: one cannot ignore the effects of radiation from AGN on their surroundings. This is perhaps an obvious point, he said, given years of accumulating evidence. The theories of AGN have developed in two directions, he said. The first in terms of understanding the central engines and the second in understanding the demographics of such systems in the universe. However, although a now well-known phenomenon, very little time has been spent on trying to understand the effects of AGN on the surrounding space.

The evolution of elliptical galaxies and spheroids of spirals display family traits that are, said Ostriker, best understood in terms of a period of star formation, followed by the loss from the system of the remaining primordial gas and the reprocessed gas from dying stars. The most probable energy source for this galactic evolution is the central black hole in each system. The black hole provides direct heating of the gas due to absorption of ultraviolet and X-ray emissions from the central AGN.

Computational studies have revealed a likely mechanism for this absorption of energy on the basis of recently determined spectra of AGN. These studies show that the gas undergoes oscillations in which it cools, falls back into the centre, accretes on to the central black hole, releases radiant energy, i.e. "outbursts", absorbs some of that energy and thus is heated, expands, and cools again and so on. During these oscillations the system appears to be either a normal elliptical galaxy, for about 99.5 percent of the time and an AGN, or quasar, for the remaining time. Ostriker made several predictions about the nature of AGN and the evolution of galaxies. First, the high temperature, high-density phases are probably too brief to be observed. On a longer timescale, the cooling period, gas bubbles with low density, high temperature should be observable. Thirdly, he suggested that the cooling shell phases are unstable to instabilities known as Rayleigh-Taylor instabilities, which results in cool gas appearing filamentary. Additional detailed spectroscopic studies, will Ostriker said, provide further evidence for the importance of heat oscillations and feedback in galactic evolution.

See-through surveys

To understand the way matter has accreted into the systems we observe throughout the universe, we need a census of all the energy-producing supermassive black holes in the universe, explained Amy Barger of the University of Wisconsin. Such a census would have to include even those objects obscured by gas and dust.

Fortunately, the Chandra and XMM X-ray observatories can effectively see through the mire and have detected obscured AGN and almost fully resolved the X-ray background into discrete sources, said Barger. She pointed out that follow-up observations with ground-based telescopes - optical, sub-millimetre, and radio telescopes - are allowing astronomers to reconstruct the history of supermassive black hole accretion from the earliest times to the present.

Only a small fraction of the energy produced in the universe is in the X-ray region, and this tends to come from supermassive black holes and so X-ray surveys could provide our best window on such black holes. Once we have accounted for the EBL (extragalactic background light), the census will be complete. There could be no other energy sources to consider.

Barger explained how results from the Chandra observatory is thus showing how supermassive black holes are seen forming from the earliest times to the present, not just during the quasar era as earlier researchers had suggested. The number of distant quasars observed in the X-ray region declines with time, an observation also seen in the visible region for quasars. However, Chandra has revealed that there are an increasing number of less luminous sources at lower redshift, in other words, not so far back in the universe's history, that are obscured in the visible region.

While these sources do not consume as much matter as the distant (and so older) quasars, there are many more of them so they produce as much light at recent times as the quasars did at early times. Barger concluded that the contents of the universe have changed from a small number of bright objects to a large number of dim ones - the active supermassive black holes in the universe have downsized!

Positive feedback

The rate of star formation in a galaxy involves an input signal which is determined by the amount of gas in the galaxy and the properties of that gas. Astrophysicists believe that there is a process that allows this input to loop around and so influence the properties of the gas thus affecting the rate of star formation. This feedback loop is at the heart of understanding AGN, explained Andrew Benson, of the University of Oxford.

Feedback is not only a phenomenon that influences the rate of star formation though Benson added. Astronomical feedback is a necessary part of the galaxy formation process, without it we would not observe the galactic systems that we do. The classic example of such a loop involves local feedback. The input signal that determines the rate of star formation in this instance, is just the mass of gas available in the galaxy from which stars will form. The more gas, the faster is star formation. As stars form, very soon afterwards, supernovae occur, and these release energy into the same galaxy, which ejects gas. Less gas means slower star formation, which in turn means fewer supernovae and less gas being ejected. Negative feedback!

Global feedback might also be applied to the star formation rate in many neighbouring galaxies when photons from new stars in one cause heating of the gas in another and slow the rate of star formation in that one and so on. Applying the concept of feedback to AGN, might also be possible, suggested Benson, the process of interest would therefore be the rate of black hole formation. If feedback applies to AGN then by definition there must be a process, an input signal, and feedback to close the loop.

Benson suggested that for feedback to be applicable there must be a phenomenon that couples star formation and AGN activity. He presented evidence that points to the existence of just such a feedback loop. In particular, the energies involved are enormous in comparison to supernovae explosions: AGN release 10 to 100 times more energy. Moreover, there are, he said, very clear correlations between galaxies and AGN that could indicate a coupling between their formation; such a coupling might well be a feedback-type coupling.

Hierarchical galaxies

Martin Haehnelt of the University of Cambridge discussed the connection between the formation of galaxies and their central supermassive black holes. He explained how the tight correlation between black hole mass and the stellar velocity dispersion of galactic bulges is strong evidence that the formation of galaxies and supermassive black holes are closely linked.

First Haehnelt asked where do we find black holes and where do we not find them? Unfortunately, the tell-tale sign of a massive black hole at the centre of a galaxy - the additional gravitational pull on the stars - can only be detected in galaxies close by. Certainly we can say, that we find black holes in all nearby galaxies which have a bulge. We are yet to find a bright galaxy with a bulge and without a black hole. However, if we study smaller and more disc- like galaxies, we see amazingly that a black hole will be present but its mass scales not with the overall mass of the galaxy but with the mass of the central bulge. Moving further down in size to globular clusters and the like our knowledge about the presence of black holes becomes less certain.

Haehnelt has reviewed the model of the joint formation of galaxies and their central supermassive black holes. In this model small galaxies form first and big galaxies form by merging of smaller galaxies. In the context of this model of a hierarchical build-up of galaxies, he has begun to describe the quantitative link between the mass of the stars in the bulge of a galaxy and the black hole at its centre.

There is a good correlation between the range of stellar velocity dispersion in galactic bulges and the black hole mass, he added. Indeed, the correlation is tight, but Haehnelt admitted, we might ultimately alight on one of many possible explanations. The tightness of the correlation is at present still something of a cosmic puzzle. In our model, he explained, black holes grow from accretion of the cold gas driven to the centre of a galaxy during enormous mergers between two galaxies. The accretion of the gas leads to the release of large amounts of energy which brighten up the nucleus of the galaxy. The brightness of such AGN means that they can be seen out to very large distances.

In reviewing sky surveys, astrophysicists see that bright AGN emitting most of their light in the optical part of the electromagnetic spectrum appear to track the accretion history of black holes rather well. However, some models that assume that most of the gas is accumulated in the black hole when the accretion process is obscured from our view by intervening dust and could only be seen with forthcoming more sensitive infra-red telescopes/satellites may need revision.

It's a gas!

Clusters of galaxies are the largest gravitationally bound objects in the universe, and contain large amounts of very hot, X-ray emitting gas. Observations of this gas give us clues about how these clusters formed and evolved. Greg Bryan of Oxford University examined recent models, both analytic and numerical, that could help provide an explanation of the structure of the thermal plasma (gas) in clusters of galaxies.

Bryan first summarized one of the key observations, which is that models including only gravity and ideal gas dynamics do not reproduce the observed relationship between a cluster's X-ray luminosity and its temperature. This is a clue that some other processes, such as heating from supernovae or AGN, are important.

Bryan explained how astrophysicists can focus on the distribution of disorder, or entropy, in a cluster and can use this quantity to better understand the evolution of the gas in clusters. This is a very useful quantity because it provides a trace of how much heating and cooling the cluster gas has been exposed to. Using a simple model, he showed that gas with an entropy lower than that observed is also gas that would be expected to cool and form stars (or accrete onto black holes) within the lifetime of the cluster. Therefore, it is not surprising that this gas should not be seen in clusters, since we would have expected it to cool out. If this model is correct, then it is primarly the physics of radiative cooling which sets the global structure of X-ray clusters.

However, continued Bryan, if the gas simply cooled and formed stars, it would form many more stars than are observed in clusters of galaxies. Therefore, something must prevent this gas from forming stars. This can be due to feedback from supernovae and AGN which generate a great deal of energy and can heat some of the cooling gas up to high temperatures and low densities. Bryan also showed the results of numerical simulations which agreed with this picture.

There are several questions that remain unanswered. For instance, what process controls feedback - stars, AGN, or something else? What happens to the gas below the cooling threshold? Does it all form stars or does some of it accrete on to the black holes at the centre of massive galaxies. Answering these questions will tell us more about how AGN (and supernovae) impact the galaxies and clusters of galaxies that we see today.

Light and sound

If energy input into galaxies from whatever source occurs at high redshift, which means further away and so further back into the history of the universe, then maybe we could observe its effects, said Andy Fabian of the University of Cambridge. He suggested that it is likely that such energy comes from powerful radio sources and focusing on a specific form of emission, so-called inverse Compton emission, from such objects, might provide us with the means to observe how it affects the intracluster medium. The space between clusters of stars in a galaxy containing hot gases.

Past energy injection from radio galaxies is likely to be responsible for changing the structure, or scaling relations, for this intracluster gas. This energy could account for structural changes that gravity alone could not. One piece of evidence that it might do so is seen in the observation that present energy injection is likely to offset much of the radiative cooling expected in many cluster cores, explained Fabian. The Chandra and XMM-Newton X-ray observations have helped shed light on these issues and point to the need for a present-day continuous gentle heat source to explain the structures we see.

However, the picture is more complex. Ripples and weak shocks seen in the X-ray pictures suggest that energy may also dissipate through these mechanical effects and so affect the overall structure. Bubbles too have been observed in galactic systems and these produce sound waves that oscillate over a large time period, some 100 million years. These sound waves create weak shocks that can lead to heating and so energy dissipation. It is inevitable that such sound waves travelling through the intracluster medium will have an effect on its structure.

Fabian suggested that there is a case to made for this intracluster medium being viscous, which would help explain how it can respond in the way it does to sound waves rather differently from the conventional picture of the medium as completely free flowing. The viscous dissipation of heat energy may, he added, turn out to be a key process.

There are several issues yet to be resolved. First, asked Fabian, can AGN heating through bubbling and sound waves be a universal solution? How does feedback, between the energy lost and the matter accreted, work over such a large region? One crucial question remains unanswered: cooling is not completely offset by the dissipation effects. Surely, cooling must also have occurred in the first place for the central temperature drop to happen, Fabian concluded.

Other articles that may be of interest:

Advanced Photon Source at Argonne National Laboratory
The growing problem of biopiracy
Grids for chemists
Deep-sea exploration - scientists under pressure
forensic science
Accidents will happen
Predicting climate change
Green silicon production
P2P for scientists
Women in science
Academic poaching of researchers
Permanent implantable contact lenses
Profile of ETH Zurich
Paradoxical ozone

Previous, previous article - The elements that make up our bodies