In most parts of astrophysics, high science priorities call for ELT data. Still, some fields will both benefit especially from ELT data and define the cutting edge of astronomy. They deal with bjects in and the outer parts of the solar system, formation of stars and planetary discs, stellar evolution, planetary systems, the evolution of galaxies, large-scale structure, the first galaxies and ionisation. Further, the regime of the unexpected carries very high weight and will most probably provide new knowledge of highest impact.
The science case defines the list of requirements for the ELT design. In all mature ELT projects, considerable work has gone into a discussion of the science case (Najita et al, 2003; Strom et al., 2003; Gilmozzi, 2004; Ardeberg, 2004; Hook, 2004; Codona, 2004). This discussion has formed the basis of the corresponding ELT design requirements.
With VLT data, the latest decade has seen a dramatically new picture and understanding of the solar system. This concerns planets, moons, comets, asteroids and, not least, objects in the Kuiper Belt. Both the light collection and the spatial resolution of an ELT will allow an even more revolutionary new chapter of such investigations.
Surface studies of distant planets and moons will provide new insights. For Io and Triton, multi-wavelength time-series work on subsurface and surface activity and atmospheric ejections and pattern evolution is of prime interest for life-fostering environments and biospheres. While space-craft fly-by missions are powerful snapshot-oriented tools, they cannot provide the important combination of spatial resolution, light collection and time-series facilities needed for a deeper insight into the processes of comets prior to, during and following close perihelion passage, sublimation and ejection. An ELT can resolve asteroids, smaller as well as larger, allowing detailed studies of their surfaces, shapes, orientations and physical properties.
While VLTs are crucial for detection and basic studies of Kuiper-Belt Objects (KBO), an ELT is required for work on their nature. Binary KBO diameters, light curves and radial velocities can be investigated (Snel et al., 2004). Masses, densities and collision data can be provided with high impact on KBO evolution. KBO shapes and rotations will allow conclusions on the evolution of the KB and solar system and its typicallity or atypicallity. For large KBOs, atmospheric studies can be made, as an ELT allows faint stars as probes.
Nucleus of Comet Halley. Giotto fly-by 1986. Largest nucleus extension is around 16 km.
We believe to know the basic processes of star and planet formation from the collapse of compact molecular clouds, via protostars, accretion and outflows to a stabilizing star orbited by a protoplanetary system and planet formation. Observations of emerging stellar and planetary systems support our general picture. However, as even VLTs with adaptive optics (AO) systems have a spatial resolution too small for definite conclusions, also the best current observing data serve for tentative conclusions only.
Star formation is normally seen as a massive process, involving many forming stars. Such a process is, we believe, initiated by stars with high mass and covered by interstellar dust. This is a scenario supported by recent observations (Nürnberger and Petr-Gotzens, 2002; Nürnberger et al, 2002; Nürnberger and Stanke, 2003; Nürnberger, 2003). Spatially resolved ELT NIR and MIR imaging and spectroscopy will be crucial for further progress, including follow-up studies of the formation of smaller stars.
Direct accretion of the most massive stars has been regarded as improbable. Instead, these stars should form through merging of intermediate-mass stars (Bonnell et al, 1998), especially in high-density regions (Bally and Zinnecker, 2005). These conclusions were questioned by Yorke and Sonnhalter (2002). Chini et al. (2004) imaged the formation of very massive stars with VLT NIR AO. Based on strong evidence of large-mass formation with associated bipolar mass outflows and even larger masses of the still active accretion disc, they found strong evidence in favour of accretion forming also very massive stars. High spatial-resolution and contrast-ratio ELT NIR data combined with dual-linear polarimetry (Vakili, 2003) should provide vastly improved insight.
Eagle Nebula, M16 dark pillar. Star formation imaged by the HST.
Orion star-formation image assembly from the HST.
Trapezium Star Formation. Gemini picture.
We have basic ideas concerning the evolution of stars, at least those of solar-like masses and compositions. However, our insight into stellar evolution is rather superficial. For stars with the largest and smallest masses and compositions significantly different from that of the sun, our insight is at best sketchy. In spite of many efforts, we are still uncertain concerning the stellar initial mass function (IMF), a fact not least affecting our understanding of stars of high-mass and near the brown-dwarf regime (Chabrier, 2003) as well as topics related to the evolution of galaxies.
More than two years after its confirmation (Domiciano de Souza et al, 2003), in June 2003, we still lack an adequate explanation of the flat shape of the B3Ve star Achernar. Combined with its apparent stability, the shape of Achernar stresses our deficient understanding of stellar physics.
Asteroseismology is excellent for stellar-evolution studies. For our insight into solar physics, it is fundamental. Highly demanding, the method has proven feasible and reliable for the Sun. For other stars, the signal-to-noise ratio is a serious limitation (Christensen-Dalsgaard, 2005). Real advance, in photometry and spectroscopy, can be made only with the light collection of an ELT.
Stars in their terminal phases, white dwarfs, neutron stars and black holes, are faint, hard to study and little known. The ELT photon collection efficiency will change the picture drastically. Allowing millisecond-resolution photometry, it permits study of micro-variability and evolution of granular and other surface features as well as accretion gas flows and rapid non-radial oscillations and flares. Monitoring of such data will provide entirely new insight into compact-object physics.
Bowshock nebula near neutron star RXJ1856.5-3754. ESO VLT image.
Exoplanet studies have advanced spectacularly since the identification of the planet orbiting 51 Peg (Mayor and Queloz, 1995). The field has expanded rapidly concerning circum-stellar discs, planet formation and giant planets. By the end of 2005, around 150 planetary systems were confirmed and direct imaging of an exoplanet (or a small brown dwarf), albeit in a rather special system, was reported (Chauvin et al., 2004).
Potentially, T Tauri stars, evolving Solar-mass stars, are excellent for studies of circum-stellar discs and planet formation. However, they are found only beyond 100 pc. Thus, they are of little real use for investigations of circum-stellar discs and smaller planets.
The processes transforming circum-stellar envelops into discs and first planets are only vaguely known (Mouillet, 2004). The small angular sizes, faintness and power-law mass distribution of the objects, the closeness to the central stars and the high speed of events make evolutionary studies difficult, even with favourable space orientation and distance from the Sun. Detailed study has been possible in a few cases only. The results indicate a large diversity. Adequate understanding of the processes leading to planet formation, not least the first phases of formation, can be achieved only with the help of ELT spatial resolution and light collection. A high frequency of circum-stellar discs around forming stars is indicated by K-band silhouette images (Elston et al., 2003). Imaging, photometry, polarimetry and spectroscopy are required, all with high spatial resolution, for studies of linear sizes, masses, morphology, chemical composition and dynamics.
Direct imaging of exoplanets with VLTs is possible. However, only high-mass planets orbiting faint stars, emerging stars and brown dwarfs, will be accessible. Imaging and, even more, spectroscopy of less massive planets accompanying other types of stars, such as solar-type stars, cannot be achieved without an ELT. While Earth-like planets are a true challenge even with ELTs, they are fully impossible targets for current telescopes.
Carina Proplyds. Images from Nathan Smith, John Bally, Jacob Thiel, Jon Morse, U. Colorado/ CTIO/NOAO/AURA/NSF.
The formation and evolution of galaxies is understood in a sketchy way only as are the results of galaxy interactions on structure and evolution. Also, our commonly accepted picture of hierarchical merging of galaxies may be in problems, as indicated by observations of galaxies around 10 Gyr old (Cimatti et al., 2004; Glazebrook et al., 2004). Our lack of insight concerns also our most nearby galaxies, in the Local Cluster (LC). Even for the relation between the Galaxy and M 31 our knowledge is weak. Classified similarly, the two galaxies show both clear similarities (Stephens et al., 2003) and dissimilarities (Brown et al., 2003).
An adequate handle on galaxy formation and evolution requires a representative galaxy sample. The LC does not provide it. The Virgo and Fornax clusters of galaxies offer excellent samples, but they are 16 and 20 Mpc away. VLTs, with full adaptive optics (AO) systems, can cover the LC but give insufficient spatial resolution at larger distances. For studies of Virgo and Fornax galaxies, an ELT is a necessity.
Stellar field populations in galaxies are hard to study (Ardeberg and Andersen, 2005; Ardeberg and Linde, 2005). Clusters offer large advantages, practically coeval and codistant and sharing a common initial heavy-element abundance. Foreground interstellar extinction can be taken as constant, internal as negligible. Clusters are easy to separate from field stars and their densities make statistics little vulnerable to field stars. They are largely unaffected by disc orientations and useful over large distances.
The weakness of stellar clusters as probes of galaxy evolution is their density, causing image crowding in distant galaxies. However, a diffraction-limited ELT keeps image crowding small enough to allow accurate photometry of single cluster stars even beyond the Virgo cluster. For younger clusters, accurate colour-magnitude diagrams (CMDs) and metallicity diagrams (MDs) can be constructed. High-quality ages can be deduced from turn-off point (TOP) phometry and metallicities from m1 versus (b-y) diagrams for galaxies beyond the Virgo cluster (Ardeberg and Linde, 2006). Even for older clusters, reliable TOP photometry can cover large distances. From integral (B-V) photometry of stellar clusters, approximate ages can be determined for very distant galaxies.
Simulated ELT star cluster Colour-Magnitude and Metallicity Diagrams at distances of 4, 8 and 16 Mpc (Ardeberg and Linde, 2006).
Studies of large-scale structure and cosmic expansion depend critically on the quality of distance reference objects. Recently, massive programmes were dedicated to important distance indicators, in the Galaxy (Kubiak and Udalski, 2003; Kervella et al., 2004), the Magellanic Clouds (Soszynski et al., 2003) and distant galaxies (Freedman et al., 2001; Saha et al., 2001; Sandage et al., 2004; Romaniello et al., 2004). Observations and theory together have much improved large-scale distance determinations. Also, more light has been shed on the dependence of distance-related parameters on metal abundance and other factors so far only partly known.
Essential concerning the calibration of distance reference objects is the science programme adopted for Gaia (Turon et al., 2005). Providing highly accurate astrometry, parallaxes, multi-band photometry and spectroscopy covering the entire Galaxy, Gaia will allow vast improvements of distance-reference calibrations. With Gaia launch in 2011, new solid reference data will be available for the ELT era. Thus, combination of ELT light collection and spatial resolution will provide a huge advance in observations of both local and distant expansion fields and cosmological parameters.
While current spectroscopy data on redshifts of galaxies reach between z = 6 and z = 7, an ELT with an adequate field of view (FoV) will, dependent on evolution, permit redshift studies of galaxies close to z = 10. The James Webb Space Telescope should operate from 2011, with identification of high-z galaxies a high-priority programme. This and new predictions of their surface distribution (Bremer et al., 2004; Bremer and Lehnert, 2005) indicate that identifications of such galaxies for ELT photometry and spectroscopy should be plentiful.
Distant galaxy Abell 1835 IR 1916 from ESO VLT images.
Results from the Wilkinson Microwave Anisotropy Probe (WMAP), the Sloan Digital Sky Survey (SDSS) and other studies guide our understanding of the evolution of the early Universe (Seife, 2003; Fan et al., 2004; Richards et al., 2004; Rojas et al., 2005; Springel et al., 2005). The picture proceeds from the first quantum variations to the fluctuations observed in the cosmic background in the microwave region. It continues around 109 years after the Big Bang with the first large structures and galaxies, and continues as the Universe evolves until the present epoch.
Still, the picture is only a first sketchy glimpse of the overall evolution. A more consistent picture requires much more detailed data on initial formation and evolution of stars, galaxies, clusters of galaxies, filaments and voids, dark matter and dark energy as well as the nature of ionisation and re-ionisation. The hydrogen, the main constituent of the Universe, had, 380 000 years after the Big Bang cooled to a neutral state. Later, it became ionised anew and has remained so.
The process of re-ionisation of the inter-galactic medium (IGM) is as important as poorly understood. While WMAP data point to a partly ionised IGM already around 500 Myr after the Big Bang (Kogut et al., 2003), observations of the earliest known quasars show that ionisation occurred first at approximately 1 Gyr after the Big Bang (Becker et al., 2001). It is unknown whether the total picture indicates two sets of sources of ionisation, a more continuous re-ionisation process or some error of interpretation.
Our guess is that IGM re-ionisation occurred due to ultraviolet radiation from pioneer stars and galaxies. However, we do not know how production of matter inhomogeneity produced the first galaxies and stars nor how the local ionisation processes of individual galaxies led to a completely ionised IGM. With VLTs, we can get a first insight into the star formation of galaxies back to around 1 Gyr after the Big Bang. At that time, in any picture discussed, the IGM was re-ionised. An adequate handle on the processes requires data on galaxies at 0.5 to 1.0 Gyr after the Big Bang. Thus, we need optical-visual and NIR ELT data, including imaging, photometry and identification spectroscopy.
Each new telescope, substantially larger and more powerful than its predecessors, has regularly exceeded expectations and satisfied the science-case requirements. Still, again regularly, its greatest contributions have been those impossible to predict, the features unexpected. With its very large leap in size and image quality, an ELT will, no doubt, reveal so far entirely unknown and unexpected processes and dimensions. It will open the door to a new cosmos and a new chapter of astronomy and physics.
Many ELT science-case programmes depend on photometry. While most interpretations from CMDs can do with a photometric precision of 0.03 to 0.06 mag, conclusions from MDs require photometry accurate to 0.01 to 0.03 mag. However, accurate photometry with an AO ELT is far from trivial. We need photometric stability both in time and over the field observed. To which precision we can achieve this is still unknown. A firm indication requires end-to-end modelling and optical path difference (OPD) data converted to point-spread functions (PSFs). Such work is currently in progress in Lund.
At first light, the ELT AO should be fully operable for wavelengths larger than 1000 - 2000 nm. However, photometry and imaging are of special interest for smaller wavelengths. This is clearly the case for abundance data, not least via photometry viable also for fainter stellar objects. Many programmes of high scientific weight require photometry in crowded (stellar) fields. For diffraction limited imaging, the resolution depends on wavelength, further emphasising observations at shorter wavelengths.
Thus, a timely extension of AO to shorter wavelengths is an item of high priority for the full scientific benefit of the ELT. Every effort should be made to secure safe AO operation also at optical-visual wavelengths a few years after first light. It is essential to have this in mind all through the development of the AO system.
For many programmes driving ELT development, (spectro-)polarimetry is essential. Star formation and circum-stellar discs are important examples as are stars distorted by non-radial pulsations and rapid rotation. Other examples are brown dwarfs, jets of active galactic nuclei and the interstellar medium.
Each ELT science-case programme defines requirements on telescope and instruments. The requirements are different and overall recommendations depend o both inclusion and priority ranking of science-case programmes. Further, it involves a balance between the requirements for first-light operation and those possible to postpone to the subsequent ELT development.
Including the science programmes discussed, the wavelength range requirement is 350 to 20 000 nm, with some extension towards longer wavelengths desired. The first-light AO range is from around 2 000 nm and upwards, with optical-visual AO a few years later. The spatial resolution required corresponds to diffraction-limited operation of a 50 m ELT or, at 2000 nm, around 10 mas. With optical-visual AO, the corresponding spatial resolution should reach around 2.5 mas. The requirements on spectral resolution range from only R = 5 (intermediate-resolution photometry) to R = 105 and beyond.
Many high-priority programmes can be carried out with an FoV of some arcsec only, while other require up to 5 and even 10 arcmin. The dilemma posed by demands on both spatial resolution and FoV is clear. Single-conjugate AO (SCAO) systems cannot cover FoVs larger than 1 arcmin and multi-conjugate AO (MCAO) systems not larger than 2 arcmin, while optional AO arrangements, such as ground-layer AO (GLAO) can cope with much larger fields, albeit unavoidably with more modest PSF performance.
Programmes on exoplanets, especially of Earth-like sizes, require contrast ratios up to and beyond 1010. Whether such contrast ratios are achievable is unclear. We note that most high-priority programmes can be conducted with contrast ratios below 107.
There are a number of additional requirements or desiderata of importance. Added to the photometric precision discussed above, examples are rapid access, time resolution, time series, long exposures, astrometry and coronagraphy. These requirements imply demands on both the telescope and its instrumentation.
So far, the science case and ELT development have progressed mainly in parallel. While requirements and performance data have been interchanged, little direct co-operation has been possible. When ELT development proceeds to a more intensive phase, improved rapports are necessary. A telescope simulation engine should produce PSFs, including time and position variations. Via end-to-end modelling, a close interaction should involve science-case, telescope and instrument teams.
Ardeberg, A. 2004, SPIE 5489, 23
Ardeberg, A. & Andersen, T. 2005, Proc. JEMAM Granada 2004, in print
Ardeberg, A. & Linde, P. 2005, Proc. JENAM Granada 2004, in print
Ardeberg, A.& Linde, P. 2006, this volume
Bally, J. & Zinnecker, H. 2005, AJ 129, 2281
Becker, R.H., Fan, X., White, R.L. & Strauss, M.A. et al. 2001, AJ 122, 2850
Bonell, I., Bate, M. & Zinnecker, H. 1998, MNRAS 298, 93
Bremer, M.N., Lehnert, M.D., Waddington, I. et al. 2004, MNRAS 347, L 7
Bremer, M. & Lehnert, M. 2005, Proc. ESO Symp. ”Exploring the Cosmic Frontier” in print
Brown, T.M., Fergusson, H.C., Smith, E., Kimble, R.A. et al. 2003, ApJ 592, L 17
Carlberg, R. 2006, this volume
Chabrier, G. 2003, PASP 115, 763
Chini, R., Hoffmeister, V., Kimeswenger, S., Nielbock et al. 2004, Nature 429, 155
Christensen-Dalsgaard, J. 2005, Proc. ESA SP-560, 13th Cool Stars Workshop, p. 81
Cimatti, A., Daddi, E., Renzini, A., Cassata, P., Vanzella, E. et al. 2004, Nature 429, 155
Codona, J.L. 2004, SPIE 5490, 379
Domiciano de Souza, A., Kervella, P., Jankov, S., Abe, L. et al. 2003, A & A 407, L 47
Elson, R.J., Lada, E.A., Alves, J., Ferreira, B., Lada, C.J. et al. 2003, AAS May 2003, 28.09
Fan, X., Hennawi, J.F., Richards, G.T., Strauss, M.A. et al. 2004, AJ 128, 515
Freedman, W.L., Madore, B.F., Gibson, B.K., Ferrarese, L. et al. 2001, ApJ 553, 47
Gilmozzi, R. 2004, SPIE 5489, 1
Glazebrook, K., Abraham, R.G., McCarthy, P.J. et al. 2004, Nature 430, 181
Hook, I. 2004, SPIE 5489, 35
Kervella, P., Bersier, D., Mourard, D., Nardetto, N., Fouqu, P. et al. 2004, A & A 428, 587
Kogut, A., Spergel, D.N., Barnes, C. et al. 2003, ApJS 148, 161
Kubiak, M. & Udalski, A. 2003, Acta Astron. 53, 117
Meinel, A.B. & Meinel, M.P. 1992, Opt. Eng. 31, 11, p. 2271
Mouillet, D. 2004, Science 303, 1982
Najita, J. & Strom, S.E. 2003, SPIE 4835, 1
Nürnberger, D.E.A. 2003, A&A 404, 255
Nürnberger, D.E.A., Bronfman, L., Yorke, H.W. & Zinnecker, H. 2002b, A & A 394, 253
Nürnberger, D.E.A. & Petr-Gotzens, M.G. 2002a, A&A 382, 537
Nürnberger, D.E.A. & Stanke, Th. 2003, A&A 400, 223
Panagia, N., Fall, S.M., Mobasher, B. et al. 2005, ApJ 633, L1
Richards, G. T., Nichol, R. C., Gray, A. G. et al. 2004, ApJS 155, 257
Rojas, R.R., Vogeley, M.S., Hoyle, F. & Brinkmann, J. 2005, ApJ 624, 571
Romaniello, M., Primas, F., Mottini, M. & Groenewegen, M. 2004, IAU Coll. 193 p. 426
Saha, A., Sandage, A., Thim, F. & Labhardt, L. et al. 2001, ApJ 551, 973
Sandage, A., Tammann, G.A. & Reindl, B. 2004, A&A 424, 43
Seife, Ch. 2003, Science 302, 2038
Snel, R., Ardeberg, A. & Flicker, R. 2004, SPIE 5382, 57
Soszynski, I., Udalski, A., Szymanski, M., Kubiak, M. et al. 2003, Acta Astron. 53, 93
Springel, V., White, S.D.M., Jenkins, A., Frenk, C.S. et al. 2005, Nature 435, 629
Stephens, A.W., Frogel, J.A., DePoy, D.L., Freedman, W. et al. 2003, AJ 125, 2473
Strom, S.E., Stepp, L.M. & Gregory, B. 2003, SPIE 4840, 116
Turon, C., O’Flaherty, K.S & Perryman, M.A.C. (eds.) 2005, ESA SP-576
Vakili, F. 2003, EAS Publ. Ser. Vol. 8, eds. Aime, C. and Soummer, R. p. 49
Yorke, H. & Sonnhalter, C. 2002, ApJ 569, 846
Wilson, R.N., Delabre, B. & Franza, F. 1994, SPIE 2199, p. 1052