Scientific Requirements for Extremely Large Telescopes (ELT's)
CapeTown (South Africa), November 14-18, 2005
This symposium, proposed by IAU Commission 9 (and supported by the
Division IX), was organised at a moment where the various ELT projects
come to the end of their phase A studies, so that a confrontation
scientific requirements with the technical possibilities seemed appropriate
before entering a more advanced phase. At the same time, it gave
opportunity to spread the information to the astronomical community
opening doors to further collaborations with the various projects.
At the occasion of the International Year of Physics (2005), the
International Union of Pure and Applied Physics (IUPAP) planned its
General Assembly in South Africa, country which also inaugurated shortly
afterwards its new giant telescope, the Southern African Large Telescope,
SALT. Holding our IAU Symposium in South Africa, jointly with those events,
was therefore an excellent opportunity to gather a wide community in a
country which has demonstrated its dynamism in
astronomical research and its constant will to participate in large
Recent examples are given by South Africa's preparation to bid to host
the SKA, its development of the Karoo Array Telescope and its Astronomy
Geographical Advantage Programme, as stressed in the opening address given
by the honourable minister of science and technology, Mr Mosibudi Mangena,
which will be reproduced in the proceedings.
This symposium was therefore sponsored jointly by IAU and IUPAP,
and most welcome initiative. As the topics of interest for ELT's
almost all of Astrophysics, it would have been difficult to accommodate
linearly in a program not to exceed one week. It was therefore decided
the SOC to start by "setting the scene" with review papers
only, and then
have the detailed discussion and analysis within four parallel session
which dealt respectively with:
A) Galaxy Formation/Evolution and Cosmology
B) Stellar Populations
C) Intergalactic medium and Large Scale Structure
D) Exoplanets and Star Formation
The last day was devoted to the presentation of the conclusions of
groups which, together with the presentation of the main characteristics
foreseen by the major projects, gave rise to a lively final discussion.
After a presentation by D. Buckley on the SALT telescope, the latest
Large Telescope (LT)
to enter into service, by our South African hosts, M. Longair reviewed
the achievements of current LT's and showed why obviously larger
telescopes were necessary. An overview of existing ELT projects by
R. Carlberg demonstrated the diversity of concepts, with well advanced
preliminary studies, and possibly their complementarity.
Exploration and detailed studies of the possible best sites over
are progressing collaboratively (M. Sarazin). K. Hodapp, in a
dynamic review of present (and future) performances of detectors,
convinced us that they will certainly meet the requirements of the
demanding instruments. On the other hand, the performances requested
adaptive optics, presented in a two-voiced show by B. Ellerbroek
and N. Hubin,
and the complexity of planned instrumentation (G. Monnet), represent
larger challenges to fully exploit the possibilities of ELT's, but
registered major progresses in recent years.
Other large projects are in the planning (SKA, presented by S. Rawlings),
are decided but will fly in some time only (JWST, discussed by J.
or will enter operations soon (Herschel, and Alma, both presented
T. Wiklind). Their impact on science to do with the ELT's can only
foreseen, but certainly calls for a fast track in erecting ELT's,
to make best use of their complementarity. In some
specific cases like the search for earth-like exoplanets, still other
techniques (like interferometry on ground or in space) might be more
efficient than direct imaging by an ELT (A. Quirrenbach),
but the latter will be indispensable for a detailed study of their
Each of the four groups studied a few scientific key-issues which
specific requirements for the design of the ELT's.
In the first splinter group (Group A, chaired by I. Hook, on galaxy
and evolution), extending for instance the relation between blackhole
and the bulge mass of their host galaxy to higher redshifts would
on the formation of blackholes, but require a high spatial resolution
sensitivity to low surface brightness, AND spatial coverage to probe
dynamical structure. Probing the dark matter in ellipticals, via
of test particles like Planetary Nebulae or Globular Clusters, requires
a wide field of view and moderate spectral resolution (R ~ 5000).
Similar requirements come from the need to study the physics of galaxies in
the redshift range from 1 to 5, but with the additional need of multiple
Integral Field Units (IFU's) to spatially resolve the objects on the kpc
scale. Studying the "First Galaxies" (at z > 6) requires both
(28th mag objects in B, or 24th in K) and high spatial resolution to allow
sampling of objects which presumably have only a few tens of milli-arc-sec
in radius, thus calling for Multi-Object-Adaptive-Optics (MOAO). However
finding such galaxies will require specific programs either with JWST or
with dedicated LT's, or perhaps ELT's themselves if wide field cameras
An interesting possibility to directly measure the cosmic expansion
(by measuring the shift with time of absorption lines in QSO's) requires
extreme accuracy ( a few cm/s/yr) and stability in velocity measurements
with high signal to noise, therefore also a high spectral resolution. The
latter is also required to constrain possible variation in fundamental
constants like the fine structure one, present measurements with LT's being
unconclusive at the moment.
In the "Stellar Populations" group (B, chaired by J. Mould),
in ELT's comes from the fact that ELT's bring galaxies within 5 MPC "as
as the Magellanic Clouds", allowing to resolve galaxies into
learning their history from "fossil records" of old stars
look-back time. Topics which were more directly discussed concerned:
Colour Magnitude Diagrams and star formation history; Galactic Archaeology
and chemical evolution; Starbursts and super star clusters; and
Neutron stars and pulsars.
Speakers in the stellar populations session offered six messages to the ELT
1) Plan for periods when adaptive optics is not operable.
2) Provide full wavelength coverage from the UV through the thermal IR.
3) Support high resolution spectroscopy.
4) Include polarization performance in the science requirements for the project.
5) Stellar populations / galaxy evolution science is strongly dependent on
6) Study the trade-off with interferometers.
The group C addressing the intergalactic medium, under chair of
divided the science in three broad topics and redshifts ranges.
At high redshift (z>7), addressing
reionisation, metals lines and Lyman-alpha emitters requires
observations at intermediate resolution
of R=2000 in the NIR with S/N up to 100. An OH-line suppressor with
multiple IFUs with field-of-view of several arcmins square is ideal.
Targets are moderately faint QSOs and Lyman-break galaxies of
AB magnitudes about 27, but require an ELT larger than 30m.
NIR observations at higher resolution (R=10000) and Signal to Noise
up to 100 would be possible with brighter, transient objects like
average-luminosity Gamma-Ray Bursts and pop.~III Supernovae.
At intermediate redshifts, to study the distribution of metals in
and their evolution with cosmic time, observation of brighter QSO's
single target mode are required, but with high spectral resolution
and high S/N (10000). To cover a large range of transitions and constrain
the ionisation corrections, the spectrograph should be blue sensitive
cover a large range in wavelength (303-930 nm).
Finally, at low redshift (z<5), the main gain of an ELT in studying
galaxies- IGM connexion is the possibility of observing fainter
targets and thus sampling the metal distribution, and its
correlation with galaxies, with a finer grid of lines along which
can be probed. The QSOs and bright LBGs (background sources for the
lines observations) can be observed
with optical, high-resolution (R=50000) spectroscopy (and S/N=100)
to probe the distribution of metals. But a detailed correlation of
metals with galaxies requires the redshift determination of the fainter
(down to 0.01 L*) using optical/NIR MOS with R=2000-5000, with
multiple IFUs and a total FoV of several arcmin-square, centered
bright LBGs and QSOs.
In group D (chaired by M. McCaughrean), Exoplanets and Star Formation
discussed, with a larger concentration on the former, revealing the
interest in this rapidly growing field. The power of an ELT at thermal-IR
wavelengths to penetrate the extreme extinction associated with massive
formation and to resolve very young high mass stars was however stressed
Zinnecker. Lenzen presented various science cases driven by thermal-IR
with an ELT, including solar system, star formation, exoplanets,
extragalactic science. Some novel instrumental concepts were presented,
including a high-precision polarimeter measuring fractional polarisation
down to one in a million (J. Hough) usable for the detection of exoplanets,
or a high-resolution IR spectrometer (R = 1 million) to detect terrestrial
planets via their OH airglow separated in velocity space from our
terrestrial airglow. It was stressed (K. Stapelfeldt) that,
based on considerable extrapolation of current AO technologies, star-planet
contrast ratios of greater than 10E8 were unlikely to be achieved
ground-based ELTs, thus making it difficult to see true terrestrial
analogues around nearby stars. Nevertheless, considerable and important
scientific results can be expected for gas and ice giant exoplanets,
as well as so-called super-terrestrials with masses roughly 10x the
In summary, strong cases were made that considerable progress can
with ELT's, with only 'modest' AO systems being required for the
Formation part, but dedicated, extremely optimised, very high Strehl
very high contrast systems for exoplanets studies, with a set of
purpose IR instrumentation for both topics.
A presentation of the actual level of preparation and advancement
studies by various countries and projects prepared the final discussion
comparison between scientific requirements and technical capabilities.
Representatives of China (X. Cui), Japan (M. Iye), Russia (Y. Balega)
presented the situation in their
respective countries. It was shown that virtual observatories and
were preparing for the ELT era also (N. Walton), while J. Urama underlined
the need of cooperation and support for astronomers in less developed
countries, to avoid too big a gap developing among various communities.
The major projects (Thirty Meter Telescope (by D. Crampton) and Giant
Magellan Telescope (by P. McCarthy) in the US, OWL in Europe (by
were presented in details, as well as the European
Design Study (presented by A. Ardeberg) joining forces between teams
Euro 50 and numerous European laboratories under the auspices of
Another two-voices presentation (US-EU, B. Ellerbroeck and C. Cunningham)
outlined the technical developments under way in the
fields of smart focal plane units, or adaptive optics, while J. Mould
addressed the needs for complementary surveys, and the various possible
operation modes in ELT's.
The final, lively, discussion adressed both technical questions like
diameter or field of view, and more general questions like time scales,
synergies with other projects or organisational matters. It was the
for the various projects to openly discuss some of their motivations,
it appeared quite clearly that the large variety of scientific projects,
and corresponding requirements should end up in a large number of
instruments which could probably not be built all at the same time
a single telescope. While it is too early to talk about specialised
consensus was that an important point is to constantly close the
interactions between scientific requirements, telescope designers
This symposium contributed to that goal, in assembling
a majority of specialists of the field, yet in a not-too-large number
to allow fruitful exchanges. It was considered by many as a
sign that the various projects had really "got off the ground",
even if the financial aspects were left aside.
Another encouraging aspect was the
participation of many enthusiastic African students in astronomy,
most of them having participated also, the week before, in an international
school of astronomy, specially organised in Cape Town by P. Whitelock
at the occasion of this symposium.
It was agreed that other events of that kind were needed in the
future, the projects developing, and the support of IAU and IUPAP,
and the IAU's role in general, were gratefully acknowledged.
Last, but not least, the conference was a success thanks also to a very fine
local organisation under the chair of John Menzies, which the delegates