THE EXTREMELY LARGE TELESCOPEIS THE ESSENTIAL NEXT STEP IN MANKINDS DIRECTOBSERVATION OF THE NATURE OF THE UNIVERSE.IT WILL PROVIDE THE DESCRIPTION OF REALITY WHICH WILL UNDERLIEOUR DEVELOPING UNDERSTANDING OF ITS NATURE.3The nuclear energy sources which provide starlightare identified, and we know that the chemicalelements of which we are made are the ash ofthat process: stardust. Exotic states of matter areknown: neutron stars, black holes, quasars,pulsars. We can show that the Universe started inan event, the Big Bang, and see the heat remnantof that origin in the cosmic microwavebackground. Tiny ripples in that background tracethe first minute inhomogeneities from which thestars and galaxies around us grew. By comparingthe weight of galaxies with the weight of all visiblematter astronomers have proven that the matter ofwhich we, the planets, the stars, and the galaxiesare made is only a tiny part of all the matterwhich exists. Most matter is some exotic stuffwhich is not yet detected directly, but has weightwhich controls the movements of stars in ourGalaxy. This dark matter or unseen mass,whatever it is made of, is five times moreabundant than are the types of matter of whichwe are made. Perhaps most exotic of all, somenew force seems to be stretching space-time,accelerating the expansion of the Universe. Thenature of this force, which controls the future ofthe Universe, remains quite unknown.Astronomy is a technology-enabled science.Progress in astronomy demands new technologiesand new facilities. Astronomical telescopes andassociated instrumentation are the essential toolswhich allow access to the widest and mostcomprehensive laboratory of all, the Universe welive in. Telescopes allow discovery of the new, andsubsequent exploration of the whole range ofknown phenomena, from Solar System objectssuch as planets, comets and asteroids, to theformation of stars and galaxies, extreme states ofmatter and space and the determination of theglobal matter-energy content of our universe.In the past half-century a new generation oftelescopes and instruments has allowed a goldenage of remarkable new discoveries.Quasars, masers, black holes, gravitational arcs,extra-solar planets, gamma ray bursts, the cosmicmicrowave background, dark matter and darkenergy have all been discovered through thedevelopment of a succession of ever larger andmore sophisticated telescopes.In the last decade, satellite observatories and thenew generation of 8- to 10-metre diameter groundbased telescopes, have created a new view of ouruniverse, a universe dominated by poorlyunderstood dark matter and a mysterious vacuumenergy density. This progress poses new and morefundamental questions, the answers to some ofwhich will perhaps unite astrophysics withelementary particle physics in a new approach tothe nature of matter. Some discoveries madeusing relatively modest technologies will requirevast increases in technology to take the next stepto direct study. Each new generation of facilities is designed to answer the questions raised by theprevious one, and yet most advance science bydiscovering the new and unexpected. As thecurrent generation of telescopes continues toprobe the universe and challenge ourunderstanding, the time has come to take the next step.In the words of the Astronomer Royal for EnglandSir Martin Rees,Cosmologists can now proclaim with confidence (but with some surprise too) that in round numbers,our universe consists of 5 percent baryons, 25 percent darkmatter, and 70 percent dark energy. It is indeed embarrassingthat 95 percent of the universe is unaccounted for: even thedark matter is of quite uncertain nature, and the dark energyis a complete mystery.A small step in telescope size will not progressthese fundamental questions. Fortunately,preliminary studies indicate that the technology toachieve a quantum leap in telescope size isfeasible. A telescope of 50-metre to 100-metrediameter can be built, and will provideastronomers with the ability to address the nextgeneration of scientific questions.ASTRONOMY IS IN ITS GOLDEN AGE.Since the invention of the telescope, astronomers have expandedmankinds intellectual horizons, moving our perception of the Earthfrom an unmoving centre of the Universe to being one of severalsmall planets around a typical small star in the outskirts of justone of billions of galaxies, all evolving in an expanding Universe inwhich planets are common.HIGHLIGHTSCIENCECASESFOR A 50METRE -100METRE EXTREMELY LARGE TELESCOPE.Are we alone? Direct detection of earth-like planets in extra-Solar Systems and a first search for bio-markers (e.g. water and oxygen) becomes feasible.Direct study of planetary systems during their formation from proto-planetary disks will become possiblefor many nearby very young stars. In mature planetary systems, detailed spectroscopic analysis ofJupiter-like planets, determining their composition and atmospheres, will be feasible. Imaging of theouter planets and asteroids in our Solar System will complement space missions.When and where did the stars now in galaxies form? Precision studies of individual stars determine agesand the distribution of the chemical elements, keys to understanding galaxy assembly and evolution.Extension of such analyses to a representative section of the universe is the next great challenge inunderstanding galaxies.Do all galaxies host monsters? Why are super-massive black holes in the nuclei of galaxies apparentlyrelated to the whole galaxy? When and how do they form and evolve? Extreme resolution and sensitivity is needed to extend studies to normal and low-mass galaxies to address these key puzzles.Can we meet the grand challenge to trace star formation back to the very first star ever formed? By discovering and analysing distant galaxies, gas clouds, and supernovae, the history of star formation,and the creation history of the chemical elements can be quantified.Were stars the first objects to form? Were the first stars the source of the ultraviolet photons which re-ionised the Universe some 200million years after the Big Bang, and made it transparent? These objects may be visible through their supernovae, or their ionization zones.Most matter is transparent, and is detectable only through its gravitational effect on moving things.By mapping the detailed growth and kinematics of galaxies out to high redshifts, we can observe darkmatter structures in the process of formation.Direct mapping of space-time, using the most distant possible tracers, is the key to defining thedominant form of energy in the universe. This is arguably the biggest single question facing physical science.In the last decades astronomy has revolutionised our knowledge of the universe, of its contents,and the nature of existence. The next big step will be remembered for discovering the unimagined new.Are there terrestrial planets orbiting other stars? Are we alone?How typical is our Solar System? What are the planetary environmentsaround other stars?When did galaxies form their stars?How many super-massive black holes exist?When and where did the stars and thechemical elements form?What were the first objects? How many types of matter exist? What is dark matter? Where is it? What is dark energy? Does it evolve? How many types are there?Extending the age of discovery...5ASTRONOMY WITH A 50METRE 100METRE TELESCOPEARE WE ALONE: HOW DO PLANETS, AND PLANETARY SYSTEMS, FORM AND EVOLVE?The science case for 50m-100m diametertelescopes is spectacular. All aspects ofastronomy, from studies of our own Solar Systemto the furthest observable objects at the edge ofthe visible Universe, will be dramatically advancedby the enormous improvements attainable incollecting area and angular resolution. Major newclasses of astronomical objects will becomeaccessible to observation for the first time. Severalexamples are outlined in the following sections.Furthermore, experience tells us that many of thenew telescopes most exciting astronomicaldiscoveries will be unexpected. Indeed the majorityof the science highlights of the first ten years ofthe first 10m telescope, the Keck, such as its partin the discovery and study of very high-redshiftyoung Lyman-break galaxies, were entirely new,violated received wisdom, and being unknown,were not featured in the list of science objectivesprior to the telescopes construction.The vast improvement in sensitivity and precisionallowed by the next step in technologicalcapabilities, from todays 6-10m telescopes to thenew generation of 50-100m telescopes withintegrated adaptive optics capability, will be thelargest such enhancement in the history oftelescopic astronomy. It is likely that the majorscientific impact of these new telescopes will bediscoveries we cannot predict, so that theirscientific legacy will also vastly exceed even thatrich return which we can predict today.In 1995 the first planet around a normal starother than the Sun was detected, by the Swissastronomers Mayor and Queloz, using a smallFrench telescope with sophisticatedinstrumentation. The rate of announcement ofnew discoveries of extra-solar planets currentlyexceeds several tens per year, with discoveriesdominated by indirect methods: either the motionof the parent star induced by the weight of theplanet, or the light-loss resulting as the planettransits in front of its star, as seen by us. Firstclaims of direct imaging of planets have alreadybeen made using 8m-10m telescopes. It is only amatter of time until several reliable discoveriesare available. Quantitative studies will becomepossible with advanced adaptive optics, usingcoronographic techniques to suppress the glarefrom the planets parent star.More detailed studies of the planets, especiallyvia spectroscopy, will however remain impossible.This is for two main reasons firstly planets arefaint; and second they are many orders ofmagnitude fainter than their parent stars, whichthey lie very close to, so that reflected light orintrinsic emission from the planet is swamped bythe glare from the star.fig.2fig.1Possibly the first ever image of an extra-solar planet around its parentstar. This adaptive optics corrected picture was taken at near-infraredwavelengths with the VLT Yepun 8-m Telescope. Right (blue): the browndwarf star 2M1 207.Left (red): a very faint companion. If indeed physically associated withthe star, this is a 5 Jupiter mass planet. The first confirmed images ofplanets will be of massive hot Jupiters, of which this may be one: anELT is essential to image planets like our Earth.778 mas55 AU at 70 pcThis 40 minutes H-band observation with the ESO VLT around a brightstar with AO correction and spectral differential imaging showsexquisite speckle subtraction. Simulated planets are added, showingthey can be detected if present. With this technique, one can detect acompanion 25,000 times fainter just 0.5 arcsec from its parent star inonly 40 minutes on an 8-m class Telescope.fake T6 planetsDeltaH = 10 mag, S/N=30at radii of 0.4, 0.6, 0.8, 1.0 arcsecExtremely Large Telescopes offer spectacularadvances in studying planetary systems, since thespatial resolution of a telescope improves inproportion to its diameter. Thus, in addition to theimproved collecting area, which is needed forobserving such faint objects as the smallerextrasolar planets, the improved resolution allowscleaner separation of a planet from the image ofits star. As a result, one of the most exciting newopportunities for Extremely Large Telescopes isthe ability directly to detect and to study largesamples of planets in other Solar Systems.Planets of course come in a wide range of types,sizes and distances from their parent stars. Whatsort of planets can be studied with different typesof telescope, and how many different planetarysystems might one be able to detect? Carefulsimulations of observations of extra-solar planetshave been completed, showing that a 30mtelescope at a standard site, equipped withsuitably sophisticated adaptive opticsinstrumentation, should be capable of studyingJupiter-like gas-giant planets out to several tens oflight years. Only a much larger, 100-m class,telescope would be capable of detecting andstudying a sample of Earth-like planets the keyhere is the extremely high spatial resolutionneeded. Earth, for example, would appear only0.1 from the sun if the Solar System wereobserved from a distance of 10 parsecs (~30 light years): see the simulation in Figure 3.The habitable zone is the narrow region in aplanetary system where water exists in liquidform. This is a pre-requisite for life as we know it.The search for earth-like planets within thatnarrow range of distances around a star requiresboth extreme light gathering power, to detect thefaint planet, and extreme telescope size, toseparate the planet from the bright star light. Thechallenge is to observe an object that is about 10billion times fainter than its parent star. Not allstars have planets, and perhaps only a very fewwill have planets in the habitable zone, so thelargest possible sample has to be surveyed if weare to be confident of identifying a true Earth-twin. The number of stars that can be studied isproportional to the spatial resolution to the cube(i.e. to D3, D telescope diameter). The time toachieve a specific reliability of detection underthese background-dominated observing conditionsis proportional to D4. A 100m telescope can inprinciple detect an Earth-like planet around asolar-type star out to a distance of 100 lightyears. This distance limit means that there areabout 1000 candidate Sun-like stars to beobserved. The corresponding numbers are about200 stars for a 50m telescope and 30 stars for a30m telescope.The large telescope collecting area, which is thekey to achieving the goal of detection of an Earth-like planet in a habitable zone, will automaticallyallow substantial extra analysis, beyond justdetection. It will also provide enough light tocharacterise planetary surfaces and atmospheres.The search for biomarkers in the planetatmosphere has the potential to discover lifebeyond our Solar System.EARTH-LIKE PLANETS AND THE SEARCH FOR EXTRA-TERRESTRIAL LIFEJupiter at 5 AUEarth at 1 AUArc seconds0.00 0.10fig.3Simulated image of a solar-type planetary system at 10 parsec (30 light-years), as seen by a 100-m telescope. This simulation subtracts thelight of the parent star, and illustrates that planetary systems like our own can be discovered and studied with the next generationtelescopes. The residual light of the masked parent star is also visible. A major challenge for the ELT is to deliver images of sufficientquality to allow this experiment.At least as important as determining the diversityof mature planetary systems is understanding theformation and early evolution processes. Is planetformation ubiquitous but survival unlikely? Orvice-versa? How long does planet formation take?How is it terminated? What happens to a planetafter it forms? All these, and many relatedquestions require detection of the observableeffects generated by on-going planet formationaround young stars. Current models, as yetuntested by direct observations, suggest thatplanets form from condensations in a dusty discencircling a young star, and subsequently createcircular gaps at discrete positions in the disc.Figure 5 shows a simulation of this process. Atelescope with sufficient resolving power and acoronagraph to suppress light from the centralyoung star will be able to detect these planetarybirthplaces, even at the inner disk locations wherehabitable planets should form. Repeatedobservation will detect directly the orbital changesin the disk as the planets grow. It is feasible tosee a dark lane from a forming Earth-like planetagainst the glare of its parent star, providedadequate contrast control is achievable.A sub-millimetre detection capability on a suitablylarge telescope would even permit the mapping ofthe colder, outer regions of proto-planetarysystems out to their Kuiper Belts, where thedebris of planetary formation is believed toaccumulate and survive.7A limitation in studies of our own Solar System isthat we have only one example. Is what we seetypical, or is it unique or transient? It is clear thata telescope and instrumentation which coulddetect Earth-like planets would with ease detectlarger planets, and planets with larger separationfrom their star. Imaging of entire planetarysystems will become possible. Such data willdefine the outcome of the formation of planetarysystems, by discovering and defining the types ofsystems which form and survive. This will tell uswhich stars have which types of planets, whatconditions are required to form the various typesof planet, what are the special properties, if any,of the parent stars and whether there are planetsaround rare types of stars , such as white dwarfsor very old halo stars.By repeated imaging, planets will be followedaround their orbits. Variations in their apparentbrightness during this process then can be usedto determine many properties. For example, theirreflectivities (albedo) determine their surfacetemperatures. For larger planets, rings like thosearound Saturn would reveal themselves (seeFigure 4). On a range of timescales, brightnesschanges resulting from diurnal rotation, weatherand even seasons offer a powerful technique forinvestigating surface and atmospheric conditions.For example, because of weather, seas, forests,deserts and ice-caps, Earths brightness changesmuch more on all time-scales than does thebrightness of Mars or of Venus.MASSIVEPLANETSANDCOMPARATIVEPLANETOLOGYWORLDSIN FORMATION9060300-30-60-90-120-150-180150120STAR0.40.30.20.10.0-180 180 -90 90 00.000 0.002 0.017 0.055 0.126 0.239 0.402 0.625Greyscale (I/F units)Full-disk albedo 0 r-4RingedNo ringPlanet orbital angle(degrees)fig.4fig.5An ELT capable of measuring contrasts of 1 part in 107to 1 part in1010could determine whether Saturn-analogues orbiting distantstars have rings or not by detecting the complicated brightness andshadow variations rings cause in the reflected starlight as afunction of time.Simulation showing the formation of gas giant planets viafragmenation of protoplanetary disks. As the planets form,detectable gaps are carved in the disk.An Extremely Large Telescope provides a naturaland valuable complement to dedicatedspacecraft. It would be capable of assembling aunique atlas of the surfaces of hundreds ofSolar System objects.Of unique value will be an Extremely LargeTelescopes ability to make repeated highly-resolved imaging and spectroscopicobservations of planets and moons withevolving surfaces and atmospheres.Detailed and continuing observations of thiskind cannot otherwise be obtained except bydedicated single-target orbiters, none of whichhave yet been sent to the outer Solar System.Figure 6 shows an image of Jupiters moon Iotaken during a fly-by of the Jupiter-orbiterGalileo. A 50m telescope would offer resolutionseveral times better than this reproduction, while a100m would be able to resolve the indicated lavaflow into more than 10 points at = 4m,measuring the temperature gradient and followingits changes as the lava cooled.Such systematic series of imaging andspectroscopic observations would, for example,resolve current issues about wind velocities andflow patterns in Neptunes atmosphere, allowstudies of changes in the contrasting terrains ofTriton, and permit the monitoring of the evolutionof the atmosphere of Pluto as it recedes from the sun.fig.6Lava flow fromPrometheusvolcanoSOLAR SYSTEM ASTRONOMYWHEN AND WHEREDID THE STARS AND CHEMICAL ELEMENTS FORM?Counting HII Regions in the Universe. The next big step in understanding theformation and histories of galaxies will follow direct resolution and analysis of theregions in distant galaxies where stars are forming, and chemical elements arebeing created and dispersed. Imaging and spectroscopy of the luminous gas cloudsassociated with the formation of massive stars is the proven technique, andrequires the sensitivity and spatial resolution of an Extremely Large Telescope.The Jovian satellite Io.fig.7When did the stars form? This basic question is akey puzzle in astronomy and is only partlyanswered. Young stars are being born today in ourown and other galaxies, but at a very low rate.Most stars were formed long ago. But when werethe legions of stars that make up the giantelliptical galaxies and the central bulges ofspirals like our own Milky Way formed? As a directmeasure we can make use of the fact thatmassive stars die young. Some stars indeed, onlya few million years after their birth die inspectacular supernovae: explosions whose flashcan outshine whole galaxies. Figure 8 illustratesone such supernova found in the Hubble DeepField (HDF) in 1997. At a redshift of 1.7 it is themost distant SN yet discovered. With anExtremely Large Telescope many such supernovaecould be seen to vast distances, corresponding toredshifts up to ten in the case of a 100mtelescope. Redshift ten corresponds to directobservation all the way back to 500 million yearsafter the Big Bang, thus seeing 97 percent of theUniverse. The frequency of supernovae at differenttimes in the history of the Universe is directlyrelated to the number of stars that formed at thatparticular cosmic epoch. Measuring the rate ofsupernova explosions across the Universe istherefore a direct way to determine when starsformed and at what rate. Simulations suggest thata 100m telescope would require about 130 nightsboth to discover ~400 supernovae using near-infrared imaging in the J, H and K bands, and tocarry out spectroscopy to confirm their nature,redshift and properties. Such a sample willprovide a reliable measure of the star-forminghistory of the Universe back to a time when theUniverse was a few percent of its present age.The more direct approach is to determine directlythe ages of stars which make up galaxies today.We can determine when stars formed by lookingat the lower-mass ones that have not blown up assupernovae, and hence remain in their host galaxyto this day. Computer models of the lives of starsshow that there is a close relation between basicobservable characteristics such as luminosity andcolour, and physical properties such as age andchemical composition.Thus, we can use observations of stars toreconstruct the ages of these objects, and hencethe times at which they formed. With a 100mtelescope it will be possible to make the requisiteobservations of stars at much larger distancesthan is possible at present. This means that wewill be able to determine the star formationhistories of entirely new classes of object, suchas elliptical galaxies. The closest examples of thisimportant type of galaxy lie well beyond thedistance that we can currently analyse, but arewithin the reach of an Extremely Large Telescope.HB-8-6-4-202M(V-I)4-1 0 1 2 3RCRGBAGBMSBLI9HDF (1995) Closeup of SN parent GalaxyDifference (97-95)Illustrating the 1997 discovery in the Hubble Deep Field of the mostdistant known supernova, at redshift z=1.7.fig.8t 0.10.1 t 0.40.4 t 1.01.0 t 3.03.0 t 6.06.0 t 10.010.0 t 13.0A predicted distribution of stars as a function of brightness (vertical axis), colour (horizontal axis) and age, fora plausible galaxy model. Stars in different age intervals are plotted in different colours, and the colour code isgiven in the figure, with units of age in billions of years.fig.9Different stellar evolutionary phases:BL - blue loop HB -Horizontal Branch RC - Red Clump RGB - Red Giant BranchAGB - Asymptotic Giant Branch MS - Main SequenceThe Hubble-Toomre sequence.The various types of galaxieswhich exist today areillustrated in the top tuning-fork diagram. The mergersequence of massiveinteractions which show howgalaxies today are mergingand changing is illustratedbelow, both in computersimulation (white dots) and inimages of real galaxies caughtin the act.fig.10To study a representative section of the Universerequires us to reach at least the nearest largegalaxy clusters which contain large ellipticalgalaxies. This means observing galaxies in theVirgo or Fornax clusters at distances of 16 or 20Mega-parsecs respectively. Simulations show thata 100m class telescope should be able to observeindividual stars in galaxies in the Virgo cluster,and determine their age and composition, evenfor the oldest, hence faintest, un-evolved stars.From these a detailed picture will be derived ofthe process by which, and the components fromwhich, the target galaxies were assembled, andthe role of dark matter in this process.To understand the creation and evolution ofgalaxies in general we must address what is oneof the major goals of future astrophysics: themapping of the distribution and growth of boththe baryonic (normal matter) and dark mattercomponents of galaxies at moderate to highredshift (z=1.5-5). Although individual starscannot be resolved at these cosmologicaldistances, the ionised gas near massive hot stars(HII regions, see figure 7) is extremely luminous,and could be detected to extremely high redshiftswith a suitable telescope. A 50m-100m ExtremelyLarge Telescope will not only resolve the distantgalaxies into their luminous components but willbe able to characterise these individualcomponents. It will win over smaller telescopes,even space-based ones, through its high angularresolution and large light collecting power. The HIIregions can be analysed to determine the relativeamounts of the elements which they contain,determining the history of the chemical elements.In addition, the velocities of these HII regionscould be measured accurately, and their velocitiesinside their parent (proto-)galaxy determined,thus allowing one to map the dark matter contentof individual galaxies throughout the observableuniverse. These individual gas clouds will then beused to trace the kinematics within the galaxiesand in their extended dark-matter haloes. Themeasurement of the kinematics of their satelliteobjects, both internal and relative to their moremassive partners, lets us estimate the amount ofmass present in the system and infer itsdistribution. This is one of the few ways we haveof detecting and examining the dark matter.This will provide us with a detailed evolutionaryhistory of the clumping of dark matterthroughout the observable universe. We will beable to obtain about one million pixel images ofvery high redshift galaxies, seeing as many detailsin such infant galaxies as we see today in nearbyones. We will see galaxy formation in all its gloryfrom formation to maturity, and so directly testour understanding of the basic evolutionaryprocesses in the Universe.11WHEN DID THE STARS ASSEMBLE INTO TODAYS GALAXIES?DO ALL GALAXY TYPES HAVE SIMILAR MERGER HISTORIES? HOW IMPORTANT IS ENVIRONMENT? HOW MANY WAYS ARE THERE TO PRODUCE AN APPARENTLY SINGLE OUTCOME?How did the galaxies that we observe around uscome to be formed? This remains one of the mostsignificant questions in modern astronomy. Thecurrent best model suggests that a hierarchicalsequence of mergers of smaller componentgalaxies built up most of the galaxies we seetoday. The merging process itself must bedominated by gravity, and so depends sensitivelyon the dominant dark matter. In consequence ofthese mergers, we should see some fossil debrisin particular stars now in the Milky Way whichoriginated as members of a now-destroyedsmaller galaxy. Indeed, recent studies of our ownMilky Way galaxy have revealed a few smallgalaxies currently merging into the Milky Way,while similar behaviour is apparent in ourneighbour the Andromeda galaxy M31. Detailedanalysis of these merger events gives clues as tothe timing of the main mergers in the Milky Wayshistory and through this clues to the role of themysterious dark matter.Until now these studies have been limited to ourown galaxy and its nearest neighbours. Muchlarger telescopes are required to study the starformation and chemical enrichment histories of arepresentative sample of galaxies.fig.11 Simulation showing stages in the formation of the galaxiesin the Local Group in a cold dark matter scenario.Snapshots are shown at various times from the earlyUniverse (z=50) to the present day (z=0).?Z=50 Z=20 Z=10Z=5 Z=1 Z=0For reasons which are not understood, theevolution and mass of super-massive black holesis apparently very closely related to the propertiesof the very much larger host galaxy. Understandingthis, and determining if it is indeed ubiquitous,would be the first clue relating the nuclei ofgalaxies to their major parts, and the first linkbetween the exotic and the typical in galaxies.How do the black holes first form? How do theygrow, and at what rate? Are growing black holesalways active? How does a central black hole knowthe properties of the larger galaxy in which itresides? Does every galaxy have a massive blackhole? Why? Progress on all of these questions requiresdetailed study of the masses and ubiquity ofcentral black holes. In the vast majority of caseswhere there is no luminous X-ray or radio sourcefor study, mapping stellar kinematics of the nucleito detect and weigh any black hole is the onlymeans available: it is in any case the most robustand reliable method. This requires the highestpossible spatial resolution and faint objectspectroscopy which is attainable only with anextremely large telescope.SUPER-MASSIVE BLACK HOLESRA offset (arcsec)Dec offset (arcsec)-0.4 -0.2 0.0 0.2-0.4-0.20.00.20.4BrH2fig.12Physical study of the environment of a billion solar mass class black hole from adaptive optics assisted spectrographic data of the centralpart (60 light years across) of the Circinus galaxy at the ESO VLT. Bottom left: ionised gas distribution (Br).Bottom Right: shocked molecular hydrogen (H2band). Br+ H2iso-velocities (km/s) showing large angular rotation around the black hole.Spectra of quasars of increasing redshiftillustrating the increase in absorption due tointervening neutral gas with increasing redshift. Atthe highest redshifts these show a Gunn-Petersonabsorption trough - the complete absorption of lightat wavelengths shortward of the Lyman - line ofatomic hydrogen - implying that the re-ionisationepoch which began at redshift ~20 must havecontinued until redshift ~6.The centres of most, perhaps all, galaxiesharbour super-massive black holes. These exoticobjects are usually discovered indirectly, asextreme radio or X-ray luminous sources, quasarsand active galactic nuclei. Direct studies, whichare critical for reliable mass determination andessential when the hole is not active, are possibleonly when precision studies of the very localregion of the galactic nucleus are feasible. Onlyrelatively close to a black hole is the gravity ofthe whole galaxy dominated by the mass of theblack hole, so that the black holes presence canbe deduced. This methodology has been provenby observations at ESO over many years whichdemonstrate the existence of a massive blackhole in the core of the Milky Way Galaxy, weighingsome 3 million times the mass of the Sun. Directmeasurements of the speed at which objects stars, gas clouds are orbiting the centre of agalaxy are required. The closer to the centre thesecan be measured, the more reliable is theevidence for the existence of the black hole andthe determination of its weight.fig.1313The early universe was hot (ionised) andtransparent. With time, the gas cooled.The aftermath of the Big Bang left the earlyuniverse an opaque gas of hydrogen and helium.Some time later the first objects heated thehydrogen and helium, making it transparent again- the era of re-ionisation. A key goal ofastrophysics is to understand how and when thefirst luminous objects in the universeformed from the primordial gas,what they were, and how theycontributed to ionising andenriching the gas withheavy elements.Tantalising questionsabout the re-ionisation historyof the universe havebeen raised byrecent results (Figures13, 14). Those from theWilkinson-MAP CosmicMicrowave Backgroundsatellite probe when combinedwith ground-based surveys of thelarge-scale structure of the Universe today,suggest that the gas was re-ionised by about180million years after the Big Bang (redshift ~ 17) while observations of the highestredshift quasars at about 700million years(redshift ~ 6) demonstrate that enough of theintergalactic medium remained un-ionised at thattime to absorb almost completely all radiationbluer than the Lyman - recombination line ofhydrogen. What is the solution to this apparentquandary? It may be that there were two re-ionisation epochs, an earlier one caused by thefirst generation of massive stars, followed bycooling, and then one later caused by the firstquasars and galaxies? Alternatively, a slower,highly inhomogeneous re-ionisation process may have occurred over the period between thetwo epochs.These models, together with other more complexpossibilities, could be tested if we can observethe ionisation state of the high-redshift universedirectly. This is feasible through analysis of theabsorption features produced in the spectra ofsuitably-luminous very distant backgroundobjects (Figure 13).There are a few populations of sources that couldclearly be observed at such very high redshift. Theshort-lived gamma-ray bursts, probably a raretype of supernova, are extremely bright for ashort time, so much so that they should bedetectable up to redshift~15-20. Normalsupernova explosions of the first stars to formwould probably be fainter than this, but could stillbe used to probe the state of the gas atredshifts up to ten. This population offirst supernovae may welldisappear once the localheavy-element enrichmentbecomes higher than1/10000 of the solarvalue. Testing thisprediction will itselfbe a majorchallenge. Quasarsare currently used aspowerful backgroundsources, and willcontinue to be useful infuture, if they exist athigher redshift. Although theepoch of first quasar formationremains an open question, the quasarsbeing found at redshifts around 6 are presumedto be powered by supermassive black holes, so weinfer that intermediate mass black holes(corresponding to quasars of intermediateluminosity) must have existed at earlier epochs,up to at least redshifts of about 10. Probing thephysics of the gas in the early Universe requiresintermediate/high resolution spectroscopy ofthese background sources in the near infrared,which is the natural domain of ground-basedExtremely Large Telescopes. Apart from the veryrare extreme gamma-ray bursters or burstscaught very early, which could be observed with a30m class telescope, spectroscopic observationsof these faint background objects can only becarried out with telescopes of the 60-100m class.THEFIRSTOBJECTS:WHERE AND WHENCAN WESEE THEM?Limits on the re-ionisation epoch of the universe. WMAP results,illustrating the residual temperature fluctuations superimposedon the cosmic microwave background radiation by the earliestgrowth of structures in the universe, imply that the epoch ofre-ionisation could have begun as early as redshift ~20.fig.14An ELT would allow study of the first galaxies,which were probably also the places of formationof the first stars, in order to resolve the questionof whether they, or the first quasars wereinstrumental in the re-ionisation of the gas in theearly universe. Candidate star-forming galaxies outto redshift about 6 have already been discoveredand a few have been confirmed spectroscopically.Equivalent objects are expected to exist out toredshifts greater than 10 for two reasons. Firstly,the analyses of the fluctuations in the cosmicmicrowave background indicate ionisation of theuniverse at redshifts >10, presumably by ultra-violet emission from the first objects. Secondly,the amount of time between redshift 10 andredshift 6.5 is so short (in cosmological terms -about 300 million years) that there is simply toolittle time to go from a universe containing nogalaxies at redshift 10 to the universe we see atredshift 6. These very high-redshift star-forminggalaxies will probably be detectable inconsiderable numbers with future spacecraft(James Webb Space Telescope) and ground-based(ALMA) facilities. However a 100m-class ExtremelyLarge Telescope will be needed to provide thedesired diagnostics of the astrophysics of boththe gaseous inter-stellar medium and the earlystellar populations in these galaxies.Furthermore, a sub-millimetre capability on anExtremely Large Telescope would allow a large-scale survey that would detect the millions ofdusty high-redshift galaxies which probablycontribute the cosmic far-infrared and sub-mmbackground, resolved down to quite faint levelsthroughout the Universe. With redshift estimatesfrom sub-mm flux ratios, such a survey wouldyield a treasure-trove of information on large-scale structure from very early epochs to therecent past.VERY HIGH REDSHIFT GALAXIESA portion of the Hubble Ultra-Deep Field showing numerous, extremelydistant, irregular (proto?) galaxies at this early cosmic epoch.fig.15Current observations indicate that dark matterdominates the dynamics of the Universe.The dark matter distribution was initially verysmooth, and became structured with time,influenced by its own gravity. This implies thatonly observations of distant, and hence faint,objects can tell us more about this growthprocess and how it really happened.The recent discovery of the accelerating expansionof the Universe has led to an urgent need tounderstand the nature of the mysterious darkenergy which is driving this expansion. The darkenergy is believed to account for about 70 percentof the energy budget of the Universe and yet itsnature is completely unknown. One potentialcandidate is the vacuum energy implied by thecosmological constant term in Einsteins fieldequations, whose solutions represent globalpictures of the Universe. However, measurementsof the effects of dark energy on cosmologicalscales constrain its contribution to be many ordersof magnitude smaller than the vacuum energyscale predicted by particle physics theories.The direct measurement of the rate at which theexpansion history of the universe has changedwith time using supernovae has shown that darkenergy exerts a negative pressure and henceaccelerates the universal expansion. Directanalysis of the expansion rates of the Universeacross space-time is needed to investigate thisremarkable form of energy. Intriguingly, most ofthe effects of dark energy are apparent atrelatively low redshifts, essentially in the timesince the Sun was formed though equivalentstudies at high redshift, when feasible, may wellhave their own surprises in store.Therefore, direct studies using well-understoodastrophysical techniques of distancedetermination are the appropriate way forward.An Extremely Large Telescope can determine theexpansion history of the universe using severaldifferent and complementary astrophysicalobjects, thus decreasing any dependency onpossibly unknown systematic effects. The well-understood primary distance calibrators,pulsating Cepheid stars, globular clusters,planetary nebulae and novae, could all in principlebe observed to distances where the effect of darkenergy first became dominant in the universe,around the time the Earth formed. The exquisitesensitivity to point sources of an Extremely LargeTelescope with appropriate adaptive opticscapability, combined with its impressive light-gathering power, will allow it to detect supernovaepossibly all the way to the time when the universefirst became transparent to light. This will allowus to map the geometry of the universe on thelargest scales.By accurately determining any variations of thestrength of dark energy with time, astronomerscan answer the fundamental question of whetherdark energy corresponds to Einsteinscosmological constant or to some quintessencefield as suggested by modern versions ofquantum field theories. The need for theseobservations is critical. The implications for all ofphysics and cosmology are vast.Cosmological observations have now become theonly way to characterise several of the mostpromising unexplored sectors beyond theStandard Model of particle physics. The discoveryand characterisation of dark matter has so farrelied entirely on astrophysical observations: suchobservations will always remain essential.The discovery and description of dark energy ispossible only with cosmological-scaleobservations: no small-scale effects are yetknown. However, dark energy and dark mattermust be part of the process of understanding thenext generation of physics theory. They arerelated to super-symmetric particles, stringtheory, theories of gravity and quantum gravity,theories of higher dimensions, and theconstancies of the fundamental constants.15DARK MATTER AND DARK ENERGYIN THE UNIVERSE:TESTINGREALITYIN ANEWDOMAINThe nature of reality: the mass-energy content of the Universe,based on the best current analyses. All normal matter is aminor perturbation. Dark matter, of unknown nature, dominatesmass. Dark energy, of unknown identities, dominates theuniverse. What is existence?10-1-2-3-4mz0 5 10 15 20 m=0 =1 m=0 =0 m=1 =0 m=0.3 =0.7groundspaceType IIType IIIPop IIIpop III plateauDetailed simulations of the types of direct indicators of distanceand past star formation which will be available to an ExtremelyLarge Telescope. Observations which are currently feasible areillustrated in blue and green. The smooth curves indicate sometypes of cosmological predictions for the evolution of Space-Time.These and other models including those not yet considered will betested directly.fig.17AtomsCold Dark MatterDark Energy73%23%4%fig.16HISTORYOF THE UNIVERSE17The star forming region RCW38. Massive young stars are forming here, illuminating the surrounding gas to make regions likethis visible across the Universe to an Extremely Large Telescope. The massive stars explode as supernovae, creating anddispersing the chemical elements, and providing probes of the history of star formation, and the geometry of space-time.fig.18The turbulent region around the ring-shaped nebula DEM L 299 in the Large Magellanic Cloud, a satellite galaxy of theMilky Way system. The Extremely Large Telescope will extend detailed studies of such regions out to high redshifts,and so quantify the creation and dispersal of the chemical elements across cosmic time.fig.1919The famous Horsehead Nebula, which is situated in the Orion molecular cloud complex. Its official name is Barnard33 and it is a dust protrusion in the southern region of the dense dust cloud Lynds 1630, on the edge of the HII regionIC 434. The distance to the region is about 1400 light-years (430 pc).fig.20Centauris A, one of the nearest massive elliptical galaxies, stillin formation as the dust lane indicates.A supermassive black hole lies at the centre of this galaxy, similar to those which power the quasars.fig.2121Protostar HH-34 in Orion. Three-colour composite of the young object Herbig-Haro 34 (HH-34),now in the protostar stage of evolution. Probing the physics of star and planet formation is a majorchallenge for the Extremely Large Telescope.fig.2223HOW AN ELT COMPLEMENTS CURRENT DEVELOPMENTS IN SPACE AND RADIO ASTRONOMYA SUBSET OF FUTURE SPACE MISSIONS AND THEIR GROUND-BASED NEEDSSPACE ASTRONOMICAL FACILITIESFor many observations a telescopes ability todetect faint sources scales as D2and the time tocarry out a given observation as D4, where D isthe primary mirror diameter. The relatively largeapertures which are affordable and technicallyfeasible for ground-based telescopes means thatthese facilities are the natural means to providemaximal light-gathering power. This sensitivity isoffset by the brightness of the background sky,and the attainable image quality. In general, bothof these are easier to provide in orbit, away fromthe earths atmosphere. The net effect is that it isnatural to provide complementarity between verylarge, relatively low-cost ground-based facilities,and special-purpose orbiting observatories. Forexample, for high-resolution spectroscopicapplications the new ground-based ExtremelyLarge Telescopes will have a natural balance inperformance with the next generation JamesWebb Space Telescope, the successor to theHubble Space Telescope.The primary motivation for the considerableexpense of space facilities is to allowobservations at wavelengths which are madeinaccessible from the ground because ofabsorption by the Earths atmosphere, especiallyin the far-Infrared, ultraviolet, X-ray and -rayregimes. The space observatories, such as theflagship X-ray facilities XMM-Newton andChandra, regularly discover sources which are toofaint in the wavelength range readily accessible tothe ground, the optical and near infra-red, to bedetected or investigated by existing telescopes.Routine images from the Hubble SpaceTelescopes Advanced Camera for Surveys revealobjects which are so faint the largest existingtelescopes are unable to acquire their spectra.Without spectroscopic information we can learnonly a limited amount about the basic nature andproperties of an astrophysical object. The adventof the James Webb Space Telescope, currentlyscheduled for launch in 2011, will increase thisimbalance. This space telescope will revealobjects an order of magnitude fainter than can bestudied in detail with existing telescopes on theground. Until the astronomical communityacquires complementary ground-based facilitieswhich are much larger than those available atpresent, the majority of future discoveries will bebeyond our spectroscopic reach and detailedunderstanding. This is a major reason whyastronomers are urgently seeking to beginconstruction of the first ground-based ExtremelyLarge Telescopes.Planned next-generation X-ray missions, such asXEUS and Constellation-X, will further increasethe need for a major enhancement in theperformance of our large optical-near infraredtelescopes if the new phenomena which theyreveal are to be understood. The Tablesummarises some future astronomical spacemissions and the complementarity they providefor ground-based Extremely Large Telescopes.Wavelength range, typeCMB map, submm sky survey1.7m optical tel. for stellar kinematics6.5m NIR/MIR telescope;imaging & multi-object spectraCoronograph and 6(4)x1.5m mid-IRinterferometer; exo-Earths imaging and spectraELT follow-up and support required(critical areas emphasised)Optical/Near-IR imaging & spectra of clustersof galaxies revealed by Sunyaev-Zeldovich effectELT exploits catalogue of solar-systems for exo-earth searchSpectroscopy and high-resolution imaging ofextremely faint sourcesComplementary approach to terrestrial planetfinding and spectroscopy Launch Date2007201120112014-2020Mission PlanckGAIAJames Webb Space TelescopeTPF and DARWINFORTHCOMINGRADIOASTRONOMY FACILITIESNew radio-astronomy ground-based facilities arebeing built that will naturally complement anExtremely Large Telescopes optical and infra-redcapabilities, and discover sources which willdemand further study at other wavelengths.The Atacama Large Millimetre Array (ALMA) willbe an interferometric array of up to 64 antennae,each 12m in diameter, located at Llano deChajnantor in the high Atacama of Chile. Thishigh, dry site will offer exceptional atmospherictransmission, permitting work at all sub-millimetre bands including the 200m windowwhich has not yet been exploited at any telescope.ALMA will provide very high sensitivity and spatialresolution beyond the limits of current ground-based telescopes. It is expected to becomeoperational early next decade and will cover a verywide range of science, detecting both thermalcontinuum emission from dust and line emission,especially from CO, in objects ranging from thenearest star-formation regions, where it will revealstructures down to ~3 AU in size, to luminousgalaxies at redshift 20. Ground-based ExtremelyLarge Telescopes will be ideally matched toprovide imaging and spectroscopic follow up ofthese sources at optical to mid-infraredwavelengths, with matched angular resolution.ALMA is under construction, and will be fullyoperational by 2012.The Low-Frequency Array (LOFAR) will comprisearound 10,000 small antennae, probably locatedacross northern Netherlands and WesternGermany. It will operate in the frequency bandsbetween 10-90 MHz and 110-300 MHz. LOFAR isdue for completion in 2008, and is a precursor toa more ambitious project, the Square KilometerArray. LOFAR will have a very wide range ofscientific applications, from ionospheric physicsto studies of the epoch of re-ionisation. AtLOFARs wavelengths most sources aresynchrotron emitters, but it will be sensitive toneutral hydrogen emissions and absorptions, atredshifts 3.5