No.

114 − December 2003

Spectrum of the star HD 37495 observed in the framework of the UVES Paranal Observatory Project (see S. Bagnulo et al., page 10)
Telescopes and Instrumentation
ALMA
G ROUNDBREAKING C EREMONY
MOMENTOUS EVENT took place on

A 6 November, at the site of the

ALMA “Operations Support
Facility” (OSF), near Chajnantor where
the array will be built. About 170 scien-
tists and dignitaries from Europe, North
America, Japan and Chile attended the
groundbreaking ceremony for this global
project. The pictures shown here and
two of the speeches that were given
(those by the ESO Director General and
President of Council) tell the story.

ALMA will be the highest-altitude, full-

time ground-based observatory in the
world, at 5,000 metres altitude. Work at
this altitude is difficult.
To help ensure the safety of the scien-
tists and engineers at ALMA, operations The President of Council, Professor P. van der Kruit, energetically breaks the ground for ALMA
will be conducted from the OSF, a com- along with Dr. W. Van Citters (NSF, Director of the Division of Astronomical Sciences) and Prof.
pound located at a more comfortable M. Tarenghi (ALMA, Director).

altitude of 2,900 metres, between the

cities of Toconao and San Pedro de Ata-
cama, and within relatively easy reach of
the array itself. The OSF will also be the
base for the construction teams, and the
64 antennas will be assembled here.
A number of containers and other facili-
ties are already located at the OSF, and
it is very reminiscent of the early days of
the construction of the VLT. Work on the
OSF and on the ALMA site itself will now
take place at an accelerated pace,
paving the way for the arrival of the first
elements of the array.

Also, at the Ground breaking ceremony,

the new ALMA logo (see facing page)
was unveiled.
Directions to everywhere from the site of the OSF, where the groundbreaking took place.

2 The Messenger 114

 ALMA Ground-breaking

A ddr ess by P ROF. P. VAN DER K RUIT ,

P r esident of ESO's Council
Mrs. Paulina Saball, Undersecretary Astronomy, astrophysics and nuclear always stressed that it was the unex-
of the Bienes Nacionales, physics have made it possible for us to pected to look forward to. So will it also
Mr. Jorge Molina, Intendente of the understand how the chemical elements be for ALMA.
Second Region, were formed. I regard this as one of the In order to build an instrument like
Distinguished Ambassadors, greatest accomplishments of science in ALMA we need a site in a very dry cli-
Esteemed Authorities, the twentieth century. It is amazing that mate, at a high altitude and with a rela-
Dear colleagues, physical science is so powerful to make tively flat area with dimensions of order
Ladies and Gentlemen, it possible for us to appreciate our ten kilometres. We have been fortunate
origin. that such a unique site is in existence
Astronomers study our roots and our here in Chile at Chajnantor.
E ARE CHILDREN OF THE UNIVERSE. relation as human beings to the cosmos. After a number of initiatives in various

W Actually, we are children of

the universe in a very strict
sense. Look at our bodies. By weight we
But astronomy is an observational
science. We will not understand the uni-
verse simply by pure thought, but rather
continents to start a project to construct
a millimetre array, eventually collabora-
tion grew out of this between North
are made up for about a quarter or so of we start by looking at it. We presently America (that is, the U.S.A. and Canada)
hydrogen. The rest is in other chemical observe in the optical with giant tele- and Europe (the member states of the
elements, of which carbon, nitrogen and scopes, such as the VLT, Gemini, Keck, European Southern Observatory and
oxygen are the major contributors. In Magellan, etc., some of which are here Spain; ESO states are Belgium, Den-
contrast, the Universe, when it was in Chile. We use telescopes in space to mark, France, Germany, Italy, the
about three minutes old and sufficiently observe at wavelengths that cannot be Netherlands, Portugal, Sweden, Switzer-
cool that atomic nuclei could exist, con- observed from the ground, such as in land and the United Kingdom). Unfortu-
sisted for three-quarters of hydrogen the X-ray region and the far infrared. We nately Japan, which was involved in the
and one quarter of helium. There was no have built very large radiotelescopes definition of the project, is not taking
carbon, no nitrogen, no oxygen or any and we have linked these or have con- part now and the original plan had to be
other chemical element except traces of structed arrays, using the same principle scaled down to what we call the “base-
lithium and boron. We now know that as ALMA will use. line ALMA”. But we are very hopeful, and
the chemical elements that make up In the last few decades, astronomers actually heard very encouraging news
most of our bodies were formed by have realized the richness of the millime- the last few days at the ALMA Board,
nuclear reactions in heavy stars that live tre and submillimetre spectrum and the that Japan will join us soon to build an
for a very short while and blow them- potential for observations there to solve even more powerful ALMA than we are
selves up as supernovae and release the the current questions in astrophysics. constructing now.
heavy elements into the interstellar gas Therefore millimetre telescopes have I would like to express, also on behalf
so that new planets and possibly life can been built, again including one on of the ALMA Board, my gratefulness to:
be formed. We are stardust. Chilean soil at La Silla, and arrays have − the visionaries who believed ALMA
been constructed in particular in North- was the biggest step astronomy could
America, Europe and Japan. These make at the present time and never gave
already use the spectral lines that can up to try to convince others;
be observed at millimetre wavelengths − the scientists and engineers that
to study the chemical composition in believed in it and showed that ALMA is
regions of star formation, especially in possible technically and financially;
the gas and dust in cool regions. − administrators and politicians that
In spite of progress in the twentieth also believed in it and convinced minis-
century, such as understanding nucleo- ters and high officials that ALMA should
synthesis mentioned above, there are be funded;
still fundamental questions left. Some − authorities that solved political and
important ones among these are the fol- legal problems;
lowing. When and how did galaxies form − and last but not least everyone at
and in what way did early star formation whatever level, in whatever capacity and
and chemical enrichment take place? from whatever country that contributed
How do planets form around young in whatever way to the fact that today we
stars? To completely solve all aspects of can formally start the construction of
these and other problems we absolutely ALMA.
need to be able to observe at millimetre
and sub-millimetre wavelengths. A los Chilenos y particularmente a la
I myself had the privilege as a gradu- gente de la comuna de San Pedro:
ate student in Leiden, the Netherlands, Muchas gracias por su cooperación en
to have professor Jan Oort as my thesis este lugar tan único y para permitir el
supervisor. This was in the days of com- desarrollo de la astronomía en su territo-
pletion of the construction of the rio hermoso. Estamos agradecidos y les
Westerbork radiotelescope and he deseamos todo lo mejor.

© ESO - December 2003 3

 ALMA Ground-breaking

A ddr ess by C ATHERINE C ES ARSKY ,

Dir ector General of ESO

T
HIS IS A GREAT DAY FOR ASTRONOMY. est technology available on this planet. We altitude, the desert, the volcanoes, the
This is indeed a great day for have come here to construct a unique population in this area, the ancient peo-
Chile, for the II Region and for instrument in these pristine surroundings, ples with their unique culture, their histo-
San Pedro de Atacama. And well aware that this vast country has a ry. There is the challenge of high techno-
this is indeed a great day for all long historical and cultural tradition of logy, the joining of so many antennas and
of us, a moment to which we have all been ancient peoples. Peoples who have asked the almost magical possibility to combine
looking forward with great anticipations. the same fundamental questions about the signals so that at the end a radio
This is the real beginning of a joint adven- the Universe and man's place in it, as we image of unequalled penetration and
ture. We will be reaching towards the now do. While the incentives and the sharpness is obtained. And then there is
stars, searching for the earliest, remotest search remain the same, we may come of course the marvelous science which
objects in the Universe, peering beyond closer to the answers with ALMA. ALMA will do, all the way from nearby
current horizons into the deep unknown. The Chajnantor plateau is a serene site stars with exoplanets in the making to
Here, on Chilean soil, in the great where man can be alone with his complex interstellar molecules and
emptiness of the Atacama desert and thoughts. It is in many ways one of the onwards to the earliest and most remote
closer to the sky than ground-based most extreme places on this planet and galaxies.
astronomers have ever been, we are now nobody who has been up there remains I sense that soon the word ALMA may
embarking upon an ambitious exploration unmoved. Once I thought of the distant also become equivalent to excitement,
of new and unknown celestial territories. past, imagining a small group of ancient, exploration of the unknown and, not least,
We do so in the service of science and daring travelers crossing that plain in front exemplary international collaboration.
society, ultimately for the benefit of of me, melting into the stark landscape. People will proudly declare that they are
humanity. They would watch the night fall, the stars part of this project. Let us rejoice that we
There have been astronomers in Chile appearing in a darkening sky, marveling at have come this far! And let us now togeth-
since long, but it was only in the early the incredible beauty of the majestic er tackle the next crucial phase with deter-
1960s that the true potential for our panorama above. Would it ever have mination. Now we begin the construction
science of this wonderful country with its occurred to them that on this very site, of this great facility in this exceptional
pure atmosphere and clear skies was hundreds of years later, a forest of giant place.
understood by scientists from North structures would be built to collect those I would like to read to you the message
America and Europe. Already in those cryptic signals from above – messages received today from Norio Kaifu, Director
early days, people from ESO and AURA from the depths of space with information of the National Astronomical Observatory
discussed opportunities to collaborate about the beginnings of that mysterious of Japan: “Congratulations for the won-
closer in their efforts to establish new and Universe in which they - and we – live? Or derful start of the ALMA construction.
powerful observatories in the IV Region. would they ever imagine that people from Breaking the ground, flying over the
However, time was not yet ready for such many other societies and from other con- Andes, the ALMA will visit a number of
joint ventures and our predecessors in the tinents would sometime assemble here, marvelous new worlds in the Universe
end decided to set up separate facilities at working together in their quest to unravel where the humankind could never reach
La Silla and Cerro Tololo. our distant origins? before it. We sincerely wish safe and suc-
ESO signed the first agreement with ALMA is indeed a unique project, both cessful construction on the Atacama site.
Chile, exactly forty years ago today. Mean- in terms of science, technology, opera- And, the third condor is ready to fly join
while, more observatories have been cre- tion, management. In addition this project you!”
ated in Chile, and in parallel Chilean possesses a great number of aspects that I express my gratitude to all those peo-
science and technology has developed fascinate young people and it provides a ple, in Europe, in North America and in
enormously. We have all benefited from fantastic opportunity to create an inviting Chile, who have helped us to reach this
increasingly closer collaboration and path towards modern science, with crucial milestone. We know that the way
many young Chilean astronomers and excitement and learning going hand in ahead is still long and that there will be
engineers are now working at these hand. problems. Together we shall solve them
observatories, also at La Silla and Paranal. Why is this so? Why has ALMA this and in not too many years we will then
ALMA is the pinnacle of this long and great appeal? There is first of all the Chaj- begin to reap the fruits of this hard labour.
steady development in which so many nantor site itself, its remoteness, the high Muchas gracias.
partners have come together to realize
what is the first truly global astronomical
project. Joining their considerable forces,
the power and experience of dedicated
specialists on three continents are now
striving to open a new, unique window
towards the Universe which will allow us
to explore vistas which have been com-
pletely hidden from view until now. We are
convinced that Chajnantor is the best
possible site for this new instrument, a
unique site which provides the ALMA tele-
scopes with optimal conditions for sensi-
tive, prolonged series of complex obser-
vations.
We are together today to celebrate the
beginnings of a great project. We are
gathering here in a beautiful and, for many With the Director General in this photograph, are (left to right)
of us, very remote region in which Robert Dickman (NSF), Eduardo Hardy (NRAO), Fred Lo (NRAO),
unspoiled nature will soon meet the high- Massimo Tarenghi (ALMA) and Daniel Hofstadt (ESO).

4 The Messenger 114

 CRIRES TAKES S HAPE
THE CRYOGENIC INFRARED ECHELLE SPECTROGRAPH BEING DEVELOPED BY ESO WILL PROVIDE A TOTALLY
NEW CAPABILITY FOR HIGH RESOLUTION (R ~ 105 ) INFRARED SPECTROSCOPY BETWEEN 1 AND 5 µm AND
OPEN UP ENTIRELY NEW FIELDS OF RESEARCH WITH THE VLT STARTING IN 2005.

A L AN M OOR WOOD * (ESO)

CRYOGENIC INFRARED ECHELLE

T
HE
Table 1. CRIRES Main Characteristics
Spectrograph (CRIRES) has had a
relatively long gestation. It was N Resolving power of ~105 from 1 to 5 µm
included in the Call for First Genera- N Adaptive optics feed to maximize SNR and
tion VLT Instruments in 1989 as one spatial resolution
of several options which were then discussed at N Echelle grating and prism pre-disperser
N Polarimetry with Fresnel rhomb retarder
the Workshop on High Resolution Spectroscopy and Wollaston Prism
with the VLT held at ESO in 1992 (High Resolu- N Pixel size ~ 0.1)
tion Spectroscopy with the VLT, ESO Conference N 0.2) u 50) slit
and Workshop Proceedings No. 40, ed. M.-H. N 4096 u 1024 InSb array mosaic
N Slit viewing camera with 1024 u 1024 InSb
Ulrich). At that time there was still a lively on- array (0.05) pixels)
going debate about the relative merits of Fourier N Calibration unit including absorption cells for
and dispersive spectrographs for high resolution accurate radial velocity measurements (< 50 m/s)
infrared spectroscopy. Pierre Connes had pio- N Limiting magnitudes ~17 (J) to 11(M) in 1 hr
neered the use of the former for very high resolu-
tion planetary spectroscopy already in the 1960s,
when infrared arrays had only 1 pixel and disper- these objectives require not only the wavelength
sive instruments could not compete with the mul- coverage but also the high spectral resolution cor-
tiplex advantage of Fourier spectrometers. By responding to a few km/s (e.g. for studying the
1992, however, low noise infrared array detectors kinematics of the cold ISM and for stellar abun-
were already available to infrared astronomers dance determinations) and with radial velocity
although still with rather small formats for echelle precision better than 50 m/s (e.g. for radial veloci-
spectroscopy. Nevertheless there was clearly a ty searches for exoplanets around cool low mass
consensus at the Workshop to include such a cryo- stars). Stellar magnetic field measurements are
genic echelle spectrograph in the instrument com- also of particular interest due to the existence of
*on behalf of the plement of the VLT. For various reasons (like fin- some particularly favourable infrared lines whose
CRIRES Team: ishing the VLT and its 1st complement of instru- magnetic Zeeman splitting will be easier to meas-
ments), the real start of CRIRES was delayed for ure than the visible lines currently used.
Peter Biereichel, several years but since its real start in 1999 the Testimony to the growing interest in high res-
Bernard Delabre, progress has been rather rapid with PDR in April olution infrared spectroscopy is the 4 day Work-
Rob Donaldson, 2000, FDR in Oct. 2001 and start of the integra- shop devoted to High Resolution Infrared Spec-
Reinhold Dorn, tion phase in Garching in 2002. troscopy which took place in Garching from 18−
Enrico Fedrigo, The characteristics of CRIRES are summa- 21 Nov. 2003. There were about 130 attendees and
rized in Table 1. Of particular interest are the fact a packed programme covering all of these topics
Gert Finger,
that the detector mosaic actually being used now as described in the summary which also appears
Gerhard Fischer,
has over 4000 pixels in the dispersion direction in this issue (see page 52).
Francis Franza, (compared with around 64 pro-
Domingo Gojak, posed in 1992) and the use of
Gotthard Huster, adaptive optics to both mini- Table 2: Science Areas
Yves Jung, mize slit losses and improve the
Hans-Ulrich Käufl, spatial resolution along the slit. N Planetary Atmospheres
Jean-Louis Lizon, N Exoplanets
N Radial velocity reflex motion searches (cool and dust
Hamid Mehrgan, SCIENCE OBJECTIVES embedded stars)
Manfred Meyer, The 1−5 µm region is rich in N Direct detection of planetary atmosphere spectral features
Jean-François Pirard, both atomic and molecular N Stars (atomic and molecular transitions, SiO, CO, CN, OH)
spectral features which offer N Abundances, COmospheres, winds, pulsations
Jérome Paufique, N Magnetic fields (Zeeman Doppler imaging)
unique probes of the chemical
Estzer Pozna, N Discs and their velocity fields
and physical conditions in a N YSO inflows/outflows
Ralf Siebenmorgen, wide range of astrophysical N Interstellar Medium (ISM) - Milky Way and nearby galaxies
Armin Silber, environments. Table 2 summa- N Chemistry and kinematics - CO,CH4,H2O,OH,H+3
Barbara Sokar, rizes some of the areas expected N Line of sight to YSOs
Jörg Stegmeier, N Extragalactic
to be of high interest for obser- N Nuclear kinematics
Stefan Wegerer vations with CRIRES. Most of N Intergalactic absorption studies

© ESO - December 2003 5

 Figure 1: Optical layout of CRIRES. Light
entering from the Nasmyth focus or the cali-
bration unit passes through the de-rotator and
adaptive optics system before entering the
cryogenically cooled spectrograph via a
dichroic window. Visible light reflected from
this window is directed to the AO wavefront
sensor. The cryogenically cooled optics com-
prises a pupil re-imaging system and Wollas-
ton prism; the pre-disperser prism and the
high resolution echelle spectrometer. The final
spectrum is imaged on a 4096 u 1024 pixel
mosaic of InSb detectors.

the prism spectrometer where it is dis-

persed and then exits through an output
slit sized to limit the wavelength range
passing into the high resolution section to
a single order. The high resolution spec-
trograph consists of a 40u20 cm, 31.6
lines/mm, 63.5 deg. blaze echelle grating
plus a TMA (three mirror anastigmat)
which acts first as the collimator and then
as the camera to image the spectrum on
the 4096 u 1024 pixel array detector.

Cryomechanical System
Figure 2 shows how CRIRES is expected
INSTRUMENT DESIGN 112, 7). The spectrograph is housed in a to look when mounted at one of the VLT
Optical Layout vacuum vessel and with its optics cooled Nasmyth focii. The main elements are the
Figure 1 shows the optical layout of to ~ 65 K and the detectors to ~ 25 K. Fol- cryogenically cooled spectrograph in its
CRIRES. Light enters from the direction lowing the input window, a pupil image is grey vacuum vessel; the table mounted
of the Nasmyth focus, either from the tel- formed at the position of a cold stop pre-optics (calibration unit, field de-rota-
escope or a calibration unit consisting of which limits parasitic background and tor, adaptive optics system coloured
an integrating sphere illuminated by con- where the Wollaston prism can be insert- green) and the electronics racks (detector
tinuum or line lamps for flat-fielding and ed. Light then either passes through the electronics in red and control electronics
wavelength calibration. Higher accuracy slit or is reflected to the slit viewing cam- in yellow). The instrument is mounted sta-
wavelength calibration is achieved using era. Light passing through the slit enters tionary on the platform primarily to
sky lines or narrow absorption lines in gas
cells which can be inserted in the beam as
shown. The gas cell slide will also contain
a specially designed ZnSe Fresnel rhomb
whose insertion can be combined with
that of a MgF2 Wollaston prism in the first
pupil image plane for measuring circular
polarization. Following the calibration
unit is a 3 mirror de-rotator which is used
to counteract the telescope field rotation
when making long slit observations. Then
comes the adaptive optics system used to
concentrate the light at the 0.2 arcsec
wide spectrograph slit. It features a 64 ele-
ment deformable mirror mounted on a
tip-tilt stage and on which is formed a
pupil image by the two mirror relay
optics. The dichroic window then trans-
mits infrared light to the cryogenically
cooled spectrograph while reflecting visi-
ble light to the wavefront sensor (WFS)
which uses an optical CCD detector and
can be translated in x,y at ~0.5 Hz to
maintain object centring as determined
by the slit viewer. As far as possible, the Figure 2: CRIRES installed at a Nasmyth focus. Shown in green are the de-rotator and part of
design of the AO system and its individual the adaptive optics sytem installed on an optical table between the Nasmyth focus and the
spectrograph. The spectrograph is cryogenically cooled and housed in the grey vacuum ves-
components have been copied from the
sel. One of the closed cycle coolers plus the turbomolecular and backing pumps are coloured
MACAO system developed by ESO for blue. The red boxes contain the front-end IRACE detector electronics and the yellow cabinets
VLTI (Arsenault et al., The Messenger, the control electronics for the complete instrument.

6 The Messenger 114

 Figure 3: One of rors are being nickel plated, diamond
the Aladdin III turned and hand post polished. The only
1024u1024 InSb
arrays mounted on non-reflecting optics in the system apart
an ESO designed from the window is the ZnSe prism used
ceramic board. for order sorting.
Cryogenic mechanisms are required
for scanning the prism (2 deg.) and
echelle grating (12 deg.), the two slits plus
the slit viewer filter and Wollaston
wheels. The scanning functions will be
driven by Phytron cryogenic stepper
motors and high precision screws and
equipped with high precision encoders.
The total mass cooled to cryogenic
temperatures is around 550 kg. Based on
our experience with ISAAC we are confi-
dent that this can be cooled down to
~65 K in 30 hours using the in-built liquid
nitrogen flow pre-cooling system. The two
closed cycle coolers are then used to
maintain the instrument at this tempera-
ture and the detectors at around 25 K.
The goal on thermal stability is to
maintain the temperature stable within
0.1 K and limit any variations of tempera-
ture gradients to < 50 mK/m/hr. As
ensure achievement of the high wave- through which the grating unit can be CRIRES is stationary, has a high thermal
length stability requirements by minimiz- accessed and removed. To the left can also inertia due to its large cryogenic mass and
ing flexure and temperature variations. be seen the pre-vacuum pump. is rather uniformly cooled by the attach-
The vacuum vessel is made of austenitic Inside, the mirror optics and most of ment of heat exchangers at several points
stainless steel with a high internal reflec- the mechanical structure is made of alu- on the cooling circuit, the short term sta-
tivity achieved by manual polishing fol- minium alloy. The TMA mirrors have a bility may be better than this. To counter
lowed by electro-polishing and inclusion thin (~30 µm) nickel coating on the drifts due e.g. to the external diurnal tem-
of a warm shield. Attached to the vessel reflective surface which has been dia- perature variations, however, active tem-
can be seen the cold head of one of the mond turned then conventionally pol- perature control is also foreseen using
two Leybold closed cycle coolers, the ished and finally ion beam polished heaters mounted on a ring whose temper-
instrument mounted turbomolecular before gold coating. Although nickel coat- ature will be controlled to ~0.1 K and is
pump, connector flanges, pressure gauges ing is usually applied on both sides we connected to various points in the instru-
and the overpressure safety valve. have found by modelling that, although ment by conducting braids. The pre-dis-
Underneath is the support and alignment reducing bending, this increases the total perser collimator mirror is also equipped
structure which also provides access to a wavefront aberration compared with plat- with piezos to allow fine active control of
port in the lower lid of the vacuum vessel ing a single surface. The remaining mir- the spectrum position using atmospheric

Figure 4: Detector
mosaic. Each of the
four array detectors
is glued to a ceram-
ic carrier of the
type shown on the
upper right and
then mounted on
the unit shown at
the upper left which
comprises a kine-
matic mount, a
copper block for
cooling and a tem-
perature sensor.
They are each con-
nected to the
acquisition elec-
tronics via 64 cryo-
genic preamplifiers
and manganin flat
band cables and
mosaiced together
as shown in the
centre.

Moorwood A., CRIRES takes shape © ESO - December 2003 7

 spectral lines for programmes requiring OPTOMECHANICS MANUFACTURE the left in the now very crowded assembly
the highest spectral stability. AND INTEGRATION hall in Garching where also SINFONI,
In order to meet the stringent thermal Most of the CRIRES hardware has now OmegaCam and various other systems
and straylight requirements the entire been manufactured and integration has are being integrated. The vacuum vessel
optical system is enclosed within a light already started in Garching. Fig. 5 shows partially visible on the far left is the old
shield plus an aluminium alloy radiation the vacuum vessel and its support struc- IRSPEC one which has been converted
shield with mirror finish quality. Care is ture on the right and the light shield on since retirement into a a cryogenic test
also being taken (e.g. by using an interme-
diate connector) to avoid light leaks at
the penetrations of cables. Essentially the
only light path into the high resolution Figure 5: CRIRES
vacuum vessel on
section of the instrument is through the the right and light
narrow order isolation slit at the exit of shield to the left in
the prism pre-disperser. the Garching
assembly hall.
Detector Mosaic
CRIRES uses 5 Raytheon 1024 u 1024
pixel InSb Aladdin arrays, one for the slit
viewer and 4 in the spectrograph focal
plane which provides a useful optical field
of 135 u 21mm. Three of the arrays are
Aladdin IIIs procured specifically for
CRIRES while the fourth is a rather old
Aladdin II, from the first best effort batch
made for ESO which yielded the ISAAC
and CONICA arrays, and which had a
number of cracks but exceptionally low
dark current (< 0.001 e-/s). These arrays
have now been re-packaged to be 3 side Figure 6: Partially
buttable so that they can be packed in a assembled cryogenic
4u1 format with a spacing between arrays pre-slit assembly.
of only 264 pixels. To do this, each array
was removed from its original LCC pack-
age by Raytheon and glued onto the spe-
cially designed ESO mount consisting of a
multilayer, co-fired, AIN (aluminium
nitride) ceramic carrier, as shown in Fig. 3,
and then glued to an adjustable invar base
plate. These are then mounted to form the
mosaic as shown in Fig. 4. The upper left
hand view shows the rear side of the
mount on which are a copper block for
the cooling braid connections, a 3-point
kinematic mount and a temperature sen-
sor. Also to be seen is the connector to the
two layer flexible manganin boards which
interface each detector to a preamplifier
board equipped with 64 cryogenic opera-
tional amplifiers. As the slit is only 512
pixels long we do not require 4 useable
quadrants per array and have thus saved
money by buying ‘reject’ arrays with one
or more defect quadrants. The re-packag-
ing has recently been completed and the
first array is now under test at ESO.
Unfortunately, the old Aladdin II, which
has a thinner substrate than the
Aladdin IIIs, was broken during the re-
packaging process and it remains to be Figure 7: Clockwise
seen if the required two quadrants are still from top left - one of the
TMA mirrors, the pre-
operational. disperser collimator mir-
The arrays will be read-out using a ror and the 40u20cm
standard ESO IRACE controller having echelle grating.
128 channels (4u32) for the science arrays
and 1u32 channels for the slit viewing
camera.

8 The Messenger 114

 fair however, some of items shown were
also delivered late and we are still await-
ing delivery from Sagem of the TMA mir-
rors which have just been accepted at
their premises in France.

ELECTRONICS AND SOFTWARE

The electronics and software for control-
ling both the CRIRES spectrometer and
the AO system are proceeding well.
Motors are being driven and their control
parameters fine tuned. The Observing
Software and Real Time Displays are
being finalized. The IRACE detector
acquisition electronics and software are
being used for the detector tests. On the
Science Operations side, the instrument
observing modes have also been defined,
the corresponding observing templates
are being coded and first thoughts are
being given to the pipeline.
Figure 8: Clockwise from top left - the TMA support structure, instrument support structure,
radiation shield and pre-slit optics housing during manufacturing.
CRIRES ON SKY?
facility used to qualify many of the principle, everything could now go ahead Unless anything goes seriously wrong
ISAAC, SOFI and now CRIRES func- full steam − were it not for those other with CRIRES or one of those other proj-
tions before integration. Figures 6 − 8 instruments just mentioned plus a few ects competing for the same manpower
show various other optical and mechani- other activities which are currently over- we expect to see 1st light on the VLT in
cal parts as described in the captions. In loading our integration laboratory. To be the first quarter of 2005. Watch this space.

ADDENDUM TO
“T HE H IS TORY AND D EVELOPMENT OF THE ESO A CTIVE O PTICS SYS TEM ”
( T HE ESO M ESSENGER N O .113)
R.N. W ILSON

F OLLOWING MY ARTICLE, l received a most interesting e-mail

from my ex-colleague of ESO and good friend, Daniel
Enard. He was concerned about the precise chronology of the
Franza and l investigated a number of possibilities of
producing S-H rasters mechanically with a German firm, but
the firm went bankrupt before any results came out. The break-
development of the ESO Shack-Hartmann image analyser − through in this procurement problem came when l gave a talk
see the second section of my article. Enard pointed out that he in Graz about Active Optics in 1981 and Franza saw a poster
and l first visited Roland Shack together in February 1976, presentation of a laser etching technique presented by the
before the 3.6 m telescope was set up some months later. RCA Laboratories in Zurich, later renamed the Paul Scherrer
Already before then I knew of the Shack proposal as I had Institute (see W99). Two successful negative masters were
been receiving the “Optical Sciences Center Newsletter”, in afterwards made to an ESO contract and replicas were made
which it was first published in 1971 (see Ref. W99 in my arti- for the final raster screens by Jobin-Yvon in Paris. This was the
cle). This was why we visited Shack to learn more about it. source of all successful screens used for the further experi-
Shack complained bitterly of lack of interest in the American ments in ESO and for the NTT. The masters for the VLT were
community and was encouraged by our deep and practical also made by the Paul Scherrer Institute, which also supplied
interest. the replicas.
In further discussions with Francis Franza, we have now Shack's original mechanical method of raster manufacture is
concluded that he and l visited Shack again in 1977 (not in now probably only of historical interest, but it nevertheless
1979 as l wrote in my article) and it was then that Shack gave showed his genius. He produced rows of cylindrical lenses by
me the lenslet raster. This was the difficult element: otherwise stepping on the lathe and fine grinding and polishing with a
the construction of the S-H image analyser was quite straight- concave cylindrical rod. Then he turned the raster through 90-
forward. It was with this original raster, made on a lathe by and produced cross-cylindrical lenses. He had proved theore-
Shack, that ANTARES l was built in 1978 (see W99, Fig. 2.24) tically that the difference between these cross-cylindrical
for testing off-line ESO telescopes on La Silla. The S-H plates lenses and true axially symmetrical lenses was well below the
were measured on the PDS measuring machine at ESO Garch- diffraction limit for such extremely weak lenses with 1 mm
ing. We believe that this was the first Shack-Hartmann image square aperture and 80 mm focal length.
analyser actually built and used for testing telescopes. In view
of its importance today in so many active and adaptive optics My thanks are due to Daniel Enard and Francis Franza for further
applications, we see this now as an important step forward in clarifying this historical development, which was of fundamental
the necessary technology. importance for both the NTT and the VLT.

© ESO - December 2003 9

 T HE UVES PAR ANAL O BSER VATORY P ROJECT
A L IBR ARY H IGH -R ESOLU TION S PECTR A OF S TARS
OF
A CROSS THE H ER TZSPRUNG -R USSELL D IA GR AM

WE PRESENT THE UVES PARANAL OBSERVATORY PROJECT, CONSISTING OF ACQUISITION, REDUCTION, AND
PUBLIC RELEASE OF HIGH-RESOLUTION, LARGE WAVELENGTH COVERAGE, AND HIGH SIGNAL-TO-NOISE RATIO
STELLAR SPECTRA OBTAINED WITH THE UVES INSTRUMENT OF THE ESO VERY LARGE TELESCOPE. DATA FOR
ABOUT 400 STARS ACROSS THE HERTZSPRUNG-RUSSELL DIAGRAM HAVE BEEN TAKEN AND REDUCED BY
MEMBERS OF THE PARANAL SCIENCE OPERATIONS TEAM AND MADE PUBLICLY AVAILABLE THROUGH A WEB
INTERFACE AT www.eso.org/uvespop.

S. B A GNULO , E. J EHIN , C. L EDOUX , R. C ABANA C , C. M ELO , R. G ILMOZZI (ESO)

AND THE ESO PAR ANAL S CIENCE O PER ATIONS T EAM

OST OF THE OBSERVA- up a library of stellar spectra with no can check thoroughly our capability of

M TIONAL CONSTRAINTS
to stellar astronomy
come from low to
medium resolution
spectra obtained in spectral bands a few
hundred nm wide or, in other cases, from
high-resolution spectra in bands a few
impact on normal operations.
Members of the Paranal Science
Operations Team have thus been author-
ized to make use of the brighter part of
twilight times for a task of public interest,
and have started the UVES Paranal
Observatory Project (UVES POP): an
reproducing stellar spectra with model
atmospheres and spectral synthesis codes,
from very normal main sequence stars, to
peculiar stars or less common objects; ele-
ment abundances can be calculated with
unprecedented accuracy for many differ-
ent kinds of star; population studies may
tens of nm wide. In spite of the capabili- on-going project of acquisition, reduction, take special advantage of the large wave-
ties of the existing instruments, there is a and public release of high-resolution, length coverage of the POP star spectra
lack of a library composed of high-resolu- large wavelength coverage, and high sig- (although it should be pointed out that
tion spectra with complete optical cover- nal-to-noise ratio (SNR) stellar spectra. UVES spectra are not absolute flux-cali-
age for stars representative of the entire Telescope time was granted under DDT brated due to slit losses); features of the
Hertzsprung-Russell (HR) diagram. This Prog. ID 266. D-5655(A). nearby interstellar medium also appear in
is due to the fact that the best instruments The targets were selected in order to great detail in the intrinsically featureless
are attached to largely over-subscribed sample the various spectral types of the spectra of the hottest observed stars. Note
telescopes, and are dedicated to very spe- HR diagram, but the limited amount of also that, although many of the observed
cific projects. At the same time, using time available during twilight (about 20 targets appear to be well-known objects,
large telescopes to observe bright targets minutes) and the desire to obtain high it is probably not unreasonable to surmise
is very inefficient, as overheads are com- SNR data, set the limiting magnitude of that serendipitous discoveries may follow
parable to, or even much longer than, the targets, as a general rule, to V=7.5. a close inspection of such a large and
open shutter times. All the observed spectra cover almost high-quality observational database –
On Cerro Paranal, twilight time is completely the wavelength interval from unknown and interesting signatures in
normally used for calibration of the 304 to 1040 nm (except for a few narrow bright stars might well have escaped
instruments in operation. At the begin- gaps). With a slit width of 0.5), the detection so far. This is particularly true
ning of each evening twilight, and at the achieved spectral resolution R = λ/∆λ is for the observed spectral region before
end of several morning twilights, tele- about 80000. For most of the spectra, the the Balmer jump (λ= 304 − 364.6 nm), that
scopes point to empty fields, in order to typical final SNR obtained in the V band represents an almost unexplored realm
provide the calibrations of the various is between 300 and 500. even for bright stars. Note also that in
instruments with fresh sky flats. At the The UVES POP library is the richest most of the cases several exposures of the
VLT Unit 2 Kueyen, the Nasmyth foci available database of observed stellar same target are taken within a very short
host two instruments dedicated to high- optical spectral lines. Many transitions period of time, with identical settings. This
resolution UV and optical spectroscopy: and oscillator strengths are often only might permit one to discover short-term
UVES and FLAMES. The UVES calibra- theoretically studied and a comparison spectral variability. Finally, it should be
tion plan is restricted to spectro-photo- with stellar spectra may permit one to pointed out that the UVES POP library
metric standard stars that are observed refine or correct the theoretical predic- may help to improve stellar spectral clas-
during the darker part of the twilights. tions. The UVES POP library is thus an sification: thanks to the large spectral cov-
With Kueyen and UVES, overheads are observational counterpart of atomic data- erage and the high resolution, one may
short enough that even the early evening bases such, as, e.g., the Vienna Atomic discover spectral indicators that have
twilight time (or late morning twilight) is Line Database (see Kupka et al. 1999 and escaped detection in previous studies.
sufficiently long to take a full spectrum of reference therein; see also [1]). The UVES POP library is also a user-
a bright star. Hence, the use of UVES dur- POP stellar spectra can be used for friendly database for teaching purposes:
ing twilights appears well suited to build many different studies. For instance, one several examples of stellar spectral fea-

10 The Messenger 114

tures may be promptly found and com- Table 1: Wavelength intervals of the UVES POP observations in DIC1 and DIC2 mode
pared among each other, in particular
thanks to a specially designed tool, the Setting Wavelength
Spectrum Preview Interface (SPI) that is Mode name coverage Optical path/arm
available at the UVES POP web site.
346B 304 – 388 nm blue
TARGET SELECTION DIC1 580L 476 – 577 nm red (lower chip)
Program stars fall into two groups, i.e., 580U 584 – 684 nm red (upper chip)
field stars, and stars belonging to two
open clusters. 437B 373 – 499 nm blue
For field stars, the only selection crite- DIC2 860L 660 – 854 nm red (lower chip)
rion applied was to cover the largest pos- 860U 866 – 1040 nm red (upper chip)
sible variety of spectral types in the HR
diagram, including peculiar objects, e.g.,
Ap and Bp stars, Wolf-Rayet stars, Be
stars, and carbon stars. Most of the logue of Ap and Am stars (Renson et al. famous stars (Aldebaran, Betelgeuse,
observed objects are brighter than 7.5 in 1991). It should be noted that the spectral Rigel, etc.). These latter stars were specif-
the V band, and in the solar neighbour- classification of the observed targets ically observed in order to allow easy
hood. Hence, most of the POP stars have came from a quick search in the litera- comparison with the observations
solar metallicity. However, a few metal ture, and that in general the spectral type obtained with other spectrographs on
poor stars (with [Fe/H] ranging from −2.5 assigned to our targets cannot be consid- smaller telescopes. The stars observed up
to −0.5) were also included in the target ered accurate or definitive. Spectral re- to the moment of writing represent about
list. Spectral types were taken from classification might eventually be per- 80% of the different spectral types and
SIMBAD and other more specialised cat- formed on the basis of UVES POP data. luminosity classes of the HR diagram. In
alogues, such as, e.g., the 14th General So far, more than 300 field stars have Fig. 1 we show the colour-magnitude dia-
Catalogue of MK Spectral Classification been observed. About 20 of them are gram of the observed field stars for which
(Buscombe 1999), and the General Cata- extremely bright (V<2.5) and very parallaxes and photometry could be
found in the HIPPARCOS catalogue
(ESA 1997).
Spectra have also been obtained for a
number of (presumed) members of two
selected open clusters, IC 2391 and NGC
6475 (M7). Due to the unforeseen move
of FORS2 from Kueyen to Yepun in P66,
the UVES service mode queues were
sometimes undersubscribed. Hence,
about 30 hours of telescope time were
used for night-time observations of some
faint objects in the two selected open
clusters (IC 2391 in P66, from 7 to 12
February 2001, and NGC 6475 in P67, in
the second half of August 2001). For tar-
get selection, extensive use has been
made of WEBDA, a WEB database for
open clusters [2]. In total about 50 stars of
IC 2391 and about 30 stars of NGC 6475
have been observed. With the open clus-
ter data, more specific scientific projects
can be carried out, for example to accu-
rately determine cluster membership,
metallicity, and angular momentum distri-
bution.
INSTRUMENT SETUP
The Ultraviolet-Visual Echelle Spectro-
graph (UVES) instrument is described in
detail by Dekker et al. (2000), in the
instrument manual by Kaufer, D’Odorico,
& Kaper (2003), and, more generally, in
[3]. UVES is a very efficient high resolu-
Figure 1: The colour-magnitude diagram of the field stars observed within the framework of the tion spectrograph designed to operate
UVES POP. (B-V) and absolute magnitudes have been computed using parallaxes and photom- from about 300 nm to 1100 nm. This
etry from the Hipparcos catalogue, and were not corrected for interstellar extinction. The
wavelength range can be almost fully
colours blue, red and black indicate an accuracy on the parallax better than 20 %, 10 % and 5
%, respectively. Stars with parallax errors larger than 20 % are not plotted. The position of the observed using two instrument modes
Sun is also indicated. (hence with two different sets of expo-

© ESO - December 2003 11

 sures), each of them characterised by the
use of a special dichroic (DIC1 and
DIC2). Dichroics are used to split the
light beam coming from the telescope to
feed the two arms of the spectrograph,
the blue arm, and the red arm. The blue
arm hosts an EEV 2Ku 4K CCD, and the
red arm hosts a mosaic of two (EEV and
MIT) 2K u 4K CCDs. Table 1 gives the
wavelength ranges covered by the UVES
POP observations.
The use of a 0.5) slit width provides a
spectral resolution of about 80 000.
Almost the full wavelength interval from
304 to 1040 nm is observed, but with a few
gaps. Two main gaps, corresponding
approximately to 577−584 nm and 854−
866 nm, are due to the physical gap
between the two chips of the red CCD
mosaic. Because of this gap, the compo-
nent at 854.2 nm of the famous CaII
triplet around 860 nm is normally not
covered in the UVES POP observations.
However, for stars with blueshifted veloc-
ities ≥ 70 km/s (prior to the conversion of
the wavelength scale to the heliocentric
rest-frame) this component may become
visible, and the one at 866.2 nm may enter
into the gap. In addition, there are several
small gaps, about 1 nm-wide each, due to
the lack of overlap between the reddest
echelle orders in the 860U setting.
The vast majority of stars were
observed with the slits oriented along the
parallactic angle, in order to minimize
losses due to the atmospheric dispersion.

DATA REDUCTION
For this project, data reduction has been a
particularly demanding task. The obvious
Figure 2: The overall spectral energy distribution of an F4V star (HD 37495) observed with
problem is related to the huge amount of UVES, together with several enlargements around the Hα absorption line. The quality of this
data that are involved, both in terms of spectrum is representative of the entire library: high SNR, high spectral resolution, large spec-
science data and calibrations. A science tral coverage; on the other hand, some ripples around 800 nm (discussed in [5]) are visible. In
the top panel, note also the gaps around 580 nm and 860 nm due to the physical gap in the
exposure in dichroic mode (i.e., when CCD mosaic of the red arm.
both arms are used) takes about
46 Mbytes (12 Mbytes for the blue arm
CCD, and 34 Mbytes for the two red arm
CCDs). In fact, each star is observed with
two instrument settings, and multiple
exposures are taken for each setting. For described in Ballester et al. (2000) (see this project, it has been found more
each observing night, and for each instru- also [4]). Together with raw data, service appropriate to use an average extraction
ment setting, five flat fields, five biases, mode UVES users receive pipeline prod- algorithm.
one wavelength calibration, plus two spe- ucts that are generally ready to be used Therefore, it was decided to take care
cial calibrations (one ‘order definition’ for science. Unfortunately, POP data can- of the data reduction with the help of a
and one ‘format check’ frames) are taken. not be processed through the standard dedicated Linux machine at ESO Vitacu-
This corresponds to 1.25 Gbytes of cali- DFO pipeline in Garching, mainly ra facilities in Chile. A number of MIDAS
bration frames. The data delivered up to because the software used by the DFO routines and display tools have been
the end of P71 (namely science and cali- group is optimised for lower SNR spectra designed for the specific format of UVES
bration data for about 400 stars) occupy (m 100). An optimum extraction algo- POP data, and for the specific require-
almost 100 DVDs! rithm is used, that assumes a Gaussian ments of the project, although the core of
UVES data obtained in service mode profile for the cross-dispersion flux distri- data reduction procedures is still the
are routinely reduced by the ESO Garch- bution. However, the use of this method UVES context v1.2 of MIDAS. An auto-
ing Data Flow Operations (DFO) group leads to poor results when applied to matic procedure has been developed to
using an automatic pipeline that is SNR M 100 data (for details see [5]). For convert the wavelength scale of the spec-

12 The Messenger 114

 QUALITY CONTROL (QC)
A number of checks were performed on
the reduced spectra, and results are
reported in a dedicated logfile that is
released together with the data. First, tar-
get coordinates written in the fits-headers
are cross-checked using the SIMBAD
database, or with specific literature, to
make sure that the target is correctly
identified. The epoch of the various
observations is also noted down, in partic-
ular it is checked that all observations of
a given star have been taken during the
same night. If not, a warning is issued in
the logfile, bearing in mind the possibility
that some stars in the sample are actually
variable.
Finally, a close inspection of the
reduced spectra is performed independ-
ently by two members of the team. In
some cases, science frames do not pass
the quality check and are rejected (for
instance, spectra with too low SNR, or
saturated, or for which extraction has
failed, or those that for any reason have a
quality substantially lower than expect-
ed). Note that even if the worst cases are
rejected, the released spectra may still be
affected by some problems like ripple
effects (due to inaccurate blaze function
correction) and lower-quality extracted
order merging. These well-known prob-
lems are documented in [5].

DATA RELEASE AND SPECTRUM

PREVIEW INTERFACE (SPI)
The data are released through an ESO
WEB page: http://www.eso.org/uvespop. The
home page gives some general informa-
Figure 3: The spectral region from 360 to 420 nm for a sample of stars of the OBAFGKM spec-
tion about the project, and points to three
tral types. All Balmer lines from Hδ down to the Balmer jump are clearly visible in the hotter main tables: i) stars belonging to the
stars. In the O star spectrum (top panel), the two sharp absorption features at about λ393.4 nm open cluster IC 2391, ii) stars belonging
and 396.8 nm are due to interstellar CaII. to the open cluster NGC 6475 (M7), and
iii) field stars ordered by spectral types. A
fourth table is devoted to some of the
brightest (V<2.5) stars in the Southern
Hemisphere.
All tables are organized in a similar
way. Each table entry corresponds to an
tra to heliocentric rest-frame, average shape of the star’s spectral energy distri- individual star, and includes links to the
multiple individual exposures (with cos- bution. SIMBAD database, to the UVES reduced
mic ray rejection), and merge the spectra The final result is one individual spec- spectra, and to a QC logfile.
from different settings all together. Dur- trum per star (plus its associated vari- For each star, the reduced spectra are
ing this process, all individual reduced ance) covering the whole optical range made available – through a direct link – in
spectra were extinction-corrected and from about 304 to 1040 nm. Some exam- the form of seven gzipped tar files, plus
flux-calibrated as explained in [6] (see ples of reduced and globally merged files one fits table. Each tar file includes a
also Hanuschik 2003) using the master are shown in Figs. 2 and 3. Figure 2 shows number of fits files that are the straight
response curves determined by ESO the large spectral coverage of UVES POP output of the MIDAS UVES pipeline
Garching Quality Control [7]. Note that observations, that are at the same time modified for this project. Six tar files are
because of slit losses occurring when a characterised by a high SNR and a high relevant, respectively, to the six observed
0.5) slit width is used under variable see- spectral resolution. Figure 3 is an example spectral ranges, i.e., 346B, 437B, 580L,
ing conditions, the final flux spectrum is of the variety of stellar spectra observed 580U, 860L, and 860U. Each of these tar
not suitable for absolute flux measure- across the main sequence of the HR dia- files includes all individual reduced spec-
ment, hence one can recover only the gram. tra and the associated variance spectra.

Bagnulo S. et al., The UVES POP © ESO - December 2003 13

 Since the merging of the various obtained out of the original raw data. S. Brillant, D. Hutsemekers, A.O. Jaunsen,
echelle orders is a delicate step of the It should be finally stated that the A. Kaufer, R. Scarpa, J.V. Smoker, T.
data reduction process, data release UVES POP reduced data are the result of Szeifert. We are also very much indebted
comes along with a seventh tar file, con- a mass production process, that cannot be to the Paranal TIOs who have been
taining the pipeline products prior to tuned to individual cases. Hence, some participating to the observations: M.T.
merging the orders, plus the sky spectra. problems will remain in the final prod- Aceveido, H. Alarcon, S. Cerda, L. Faun-
Unmerged spectra may be used to check ucts. For individual stars, better results dez, C. Herrera, N. Hurtado, A. Lopez, C.
if some features are due to an artifact pro- may possibly be obtained by retrieving McKinstry, J. Navarrete, H. Nun ~ez, A.
duced by the merging, and in some cases, the raw data and corresponding calibra- Pino, A. Ramirez, and A. Zarate.
may be preferred to merged spectra for tions at the ESO UVES archive [8] using
science use. program ID 266. D-5655(A). REFERENCES
The naming convention is self- A collection of basic cgi plotting rou- Ballester, P., Modigliani, A., Boitquin, O., Cris-
explanatory. All file names explicitly refer tines using PGPERL and CFITSIO perl tiani, S., Hanuschik, R., Kaufer, A., & Wolf,
to the star’s name (generally the entry in allows the user to overview the merged S., 2000, The UVES Data Reduction
Pipeline, The ESO Messenger, 101, 31
the Henry Draper Catalogue), the instru- spectrum of a given star in the desired Buscombe, W., 1999, 14th General Catalogue of
ment setting, and the time-stamp, when- spectral range. These routines form the MK Spectral Classification, VizieR On-line
ever applicable. For instance, a file called Spectrum Preview Interface (SPI), a dis- Data Catalog: III/222. Originally published
in: Northwestern Univ., Evanston, Illinois
hd123456_437B_2001_03_21:01:15:00.fits play tool that has been specifically creat- ISBN 0-939160-12-9 (1999)
corresponds to the spectrum of star HD ed for the public release of UVES POP Dekker, H., D’Odorico, S., Kaufer, A., Delabre,
123456 obtained using the 437B setting spectra. Users can choose to have the B., Kotzlowski, H., 2000, Design, construc-
and observed on the 21st of March 2001, spectrum displayed in GIF format, or to tion and performance of UVES, the echelle
spectrograph for the UT2 Kueyen telescope
at 01:15:00 UT (date-stamp in fact corre- save it as a postscript file. A third possibil- at the ESO Paranal Observatory, SPIE
sponds to the time of shutter opening). ity allows the user to download the spec- Munich [4008-534]
Ideally, the very final product is the trum as an ASCII file in a given wave- ESA, 1997, The Hipparcos Catalogue, ESA SP-
1200
fits table that is obtained by averaging/ length interval.
Hanuschik, R. W., 2003, A&A, 407, 1157
merging all final pipeline products as The SPI can be activated in two differ- Kaufer, A., D’Odorico, S., & Kaper, L., 2003,
explained before. This table includes the ent ways: i) through a special link from UVES User manual, Doc. No. VLT-MAN-
star’s spectrum and its variance. It is avail- the table with the target list, and ii) ESO-13200-1825
Kupka, F., Piskunov, N. E., Ryabchikova, T. A.,
able through a direct link from the WEB through the main SPI web page. The for- Stempels, H. C., & Weiss, W. W., 1999,
tables, and its name only refers to the tar- mer is the most basic and probably the A&AS, 138, 119
get name (e.g., hd123456.tfits). This file most convenient method to select and Renson, P., Gerbaldi. M., & Catalano, F. A.,
should be used with some caution. For display individual spectra. The latter 1991, A&AS, 89, 429
instance, it may come from spectra taken method permits one to perform elemen-
at distant epochs, and, in the case of vari- tary searches through the POP database
able stars, different spectral regions may (e.g., for a given name string and/or spec- WEB LINKS CITED IN THIS PAPER
[1] The VALD database:
have been produced under different phys- tral type and/or spectral peculiarity) and http://www.astro.univie.ac.at/~vald/
ical conditions. Moreover, the globally to display up to five spectra altogether. [2] The WEBDA database:
merged file may come from original spec- Overplotting has two options: spectra can http://obswww.unige.ch/webda/
tra of uneven quality in terms of, for be stacked together, or shifted along the y [3] The UVES web pages:
instance, blaze function correction. Users direction. http://www.eso.org/instruments/uves/
should check the QC logfile and the fits [4] The UVES pipeline:
http://www.eso.org/observing/dfo/quality/UVES/
headers of individual spectra. Note also ACKNOWLEDGEMENTS pipeline/
that, since the globally merged spectra We are grateful to A. Modigliani for use- [5] Problems with the pipeline reduced spectra:
have their wavelength scale converted to ful discussions about the UVES pipeline, http://www.eso.org/observing/dfo/quality/UVES/
the heliocentric rest-frame, telluric lines, D. Silva and S. Hubrig for their comments pipeline/pipe_reduc.html#problems
that in the observed frame are fixed in to the manuscript, V. Andretta, J.D. Land- [6] How to flux calibrate UVES data:
wavelength, appear mis-aligned. Telluric street, P. Prugniel for their comments to http://www.eso.org/observing/dfo/quality/UVES/qc/
lines appear at the correct wavelengths the UVES POP web pages. This project response.html
[7] The UVES master response Curves:
only in the spectra included in the has made use of the SIMBAD database,
http://www.eso.org/observing/dfo/quality/UVES/qc/
gzipped tar files. However, in many cases, operated at CDS, Strasbourg, France. std_qc1.html#response
and for many applications, the fits table is Among SciOps team members we [8] The ESO UVES archive:
close to the best product that can be acknowledge a special support from: http://archive.eso.org/wdb/wdb/eso/uves/form

14 The Messenger 114

 P USHING T ECHNOLOGIES : ESO F IBRE
L ASER D EVELOPMENT FOR L ASER
G UIDE -S TAR A D APTIVE O PTICS
IN THE CONTEXT OF DEVELOPMENTS TOWARD MATURE ADAPTIVE OPTICS SYSTEMS FOR LARGE TELESCOPES, WE
DESCRIBE THE PRESENT ACTIVITIES AT ESO IN THE AREA OF FIBRE LASERS.

D. B ONA CCINI 1 , W. H A CKENBERG 1 , L. TAYLOR 1,2

1EUROPEANSOUTHERN OBSERVATORY, GARCHING, GERMANY
2UNIVERSITY OF CALIFORNIA, LAWRENCE LIVERMORE NATIONAL LABORATORY, USA

HE NEW TECHNOLOGIES for with four more to be installed in the com- scattering of laser beams, allow the user

T astronomical Adaptive Optics

are progressively becoming
mature. Adaptive Optics in
Astronomy allows the tele-
scopes to deliver diffraction limited
images, correcting the atmospheric turbu-
lence effects: image motion and image
ing years. to point almost anywhere in the sky,
The Adaptive Optics servo control obtaining extremely large sky coverage.
needs a bright reference source in order This opens the door to AO for extragalac-
to perform the real-time correction at fre- tic astrophysics, and allows overcoming
quencies up to 1 kHz refresh rates. the first of the two limitations.
Besides, the optical wavefront correction Lick Observatory (USA), with a pro-
is valid only within isoplanatic patch totype system, has reported (Gavel et al.,
blur. ESO already has three facility Adap- areas around the reference source, which 2003) on the 3 m Shane telescope Strehl
tive Optics systems deployed on the VLT, at K-band (2.2 µm wavelength) at VLT Ratios as high as 0.6 when imaging at
sites have about 30) radius in median see- wavelengths around 2.2 µm [Fig.2]. It is
ing conditions. clear from their experience and from the
Therefore the current two limitations European experience obtained with
of Adaptive Optics are the need of a ALFA at the German Calar Alto obser-
bright reference point source, and the lim- vatory (Eckart et al., 2000), that an LGS-
ited field of view. One would wish to AO system requires multidisciplinary
extend the corrected field of view to one technologies, very well engineered sys-
arcmin radius. Adaptive Optics has in tems and an overall tuning together with
recent years successfully demonstrated the science instrument, in order to per-
on the field its capabilities with single nat- form as expected at a good astronomical
ural or laser reference sources (Takami et site.
al., 2003; Wizinowich et al., 2003). ESO To extend the field of view, the Multi-
with its community has been one of the conjugate AO technique has been pro-
first players (Beuzit et al. 1997) with facil- posed (Beckers, 1988) and is being active-
ity systems [Fig.1]. ly pursued worldwide. Multiple guide
Adaptive Optics (AO) is still a young stars of sufficient brightness and correct
and rapidly evolving technology which in separation are necessary for this tech-
its maturity promises to overcome the nique. Again this may be obtained with
current limitations. Toward this goal, use- groups of bright Natural Guide Stars
ful for 8−10 m class tele-
scopes, and vital for Extrem-
ely Large Telescopes of the
future, strategic R&D work
is being done at ESO as well
as in other observatories
around the world.
The stars are the natural
reference sources for Adap-
tive Optics; however the
brightness required limits to
less than 1% the area of the
Figure 1: K band images obtained at the sky which can effectively be
ESO La Silla 3.6m telescope with the Adonis used with Natural Guide
system, in May 1997. The Adaptive Optics Stars Adaptive Optics (NGS- Figure 2: Lick Observatory 3 m telescope results with their
closed loop (top) and open loop (bottom)
images show the difference between diffrac- AO) on 8−10 m class tele- experimental LGS-AO. The K-band Strehl ratio vs Fried
tion limits and seeing. The 5 second scopes. Laser Guide Stars parameter size, ro, is shown. A Fried parameter of 15 cm
exposure image is in log scale to show the (LGS), i.e. artificial reference at 0.5 µm wavelength corresponds to a seeing of 0.7 arc-
Point Spread Function details. sec (Courtesy of Don Gavel, Lick Obs.).
sources created by the back-

© ESO - December 2003 15

 which happen to be of the right magni- We report briefly in this article on the Therefore the high power laser beam
tude and angular separation, the so-called fibre laser program at ESO, which has quality has to be also quite good (better
“asterisms”, or with multiple Laser Guide provided so far two Patents for our organ- than 1.3 times the diffraction limit beam
Stars (Ellerbroek et al., 2003). The latter isation. We will report on the fibre relays Full Width Half Maximum, M2 = 1.3).
give a preciously uniform Point Spread in a coming article. On the other end, the different LGS
Function across the field of view in closed ESO has carried out internally, and beams do not have to be coherent and
loop. then double checked with consulting may be generated separately or the lasers
Around the world there are at the companies, the fibre laser design and its distributed at different locations. Other
moment more than 13 telescopes being numerical simulations. We are now in the requirements apply specifically to the tel-
equipped with single or multiple LGS- breadboarding phase of a prototype fibre escope environments. The laser should be
AO systems. laser to be transferred later to European efficient in term of power conversion (i.e.,
ESO is integrating a Laser Guide Star industry for commercialization. less energy to dissipate per watt emitted).
Facility (LGSF) on the VLT [see Messen- It should be rugged, turn-key and ideally
ger No. 100], which will serve the NACO require little or no maintenance.
and SINFONI instruments equipped with WHICH LASERS Off the shelf solid state lasers at
AO. We are building the LGSF with pro- Laser guide stars created in the mesos- 589 nm with these characteristics do not
vision for up to 5 LGS to be projected and pheric sodium layer at an altitude of yet exist.
diagnosed by the system [Fig 3], in order about 90 km (and a wavelength of The suitable 589 nm lasers available
to be ready for an upgrade toward Multi- 589 nm) play a major role, since they are today are only of the dye type. One model
conjugate Laser Guide Star Adaptive generated at the greatest possible dis- has been built by Lawrence Livermore
Optics. tance in the earth’s atmosphere, making National Laboratory for Lick Observato-
Although much research is still in them most star-like and minimizing focus ry and Keck, a different model is built by
progress, looking at the past trend it is anisoplanatism effects. the Max Planck Institut für Extrater-
likely that in the timeframe of 10 years The requirements for a sodium-guide restrische Physik in Garching as part of
LGS-MCAO will become a mature, star laser for multiple-guide-star AO sys- the ESO-LGSF collaborative project.
advanced, deployable technology enabl- tems are stringent: it needs to be an effi- Very recently the US Air Force has
ing AO to overcome or make negligible, cient, highly reliable and turn-key system. demonstrated a sum frequency laser able
the current limitations of guide star The output wavelength needs to be pre- to produce 20 W, which will be scaled up
brightness and field of view. By the use of cisely centred on the mesospheric, 3 GHz to 50 W CW. Development programs for
single or multiple laser guide stars current wide sodium D2 line at 589 nm. The for- discrete solid-state lasers are in place at
and future adaptive optics (AO) systems mat may be Continuous Wave (CW) or GEMINI Observatory and the University
will provide imaging at the diffraction pulsed (800 Hz repetition rate, 200 of Chicago.
limit of resolution over a wider field of microsec duration pulses) to exploit tech- For the next-generation laser guide
view and at shorter wavelengths than cur- niques which reduce or eliminate the per- star and multiple guide-star adaptive
rently possible. This is because laser guide spective elongation effects. The equiva- optics systems ESO’s strategy is to devel-
stars (LGSs) can be virtually projected lent power to be emitted is 10 W (15 W op 10 W fibre lasers at 589 nm, together
anywhere in the sky and made bright goal) per LGS. A polarised output is with industry. This strategy has created a
enough to fulfil the utmost flux require- desirable for optical pumping of the collaboration agreement with the
ments for real-time compensation of the mesospheric sodium atoms in order to Lawrence Livermore National Laborato-
atmospheric turbulence. enhance the resonant backscatter signal. ry (LLNL) in the U.S to develop a sum-
The ESO group working on the Laser For multiple sodium guide-star adaptive frequency fibre laser and an internal
Guide Star Facility project, aimed at LGS optics at least four laser guide stars will effort on a less complex fibre Raman
Adaptive Optics, is following this strate- be required on an 8 m class telescope. laser. The decision for fibre lasers was
gic view. For this reason it is pushing The laser beam quality when emitted based on a number of advantages which
R&D for two different technologies cru- at the launch telescope has to have a are hard to overestimate. Moreover they
cial to LGS-AO systems for astronomy: wavefront variance equal or comparable have been very recently demonstrated
the fibre lasers and the high power laser with the variance introduced by the with powers up to 500 W at visible wave-
beam relays based on Photonic Crystal atmospheric turbulence in the uplink lengths, both in CW and pulsed formats.
Fibres. beam (from telescope to Mesosphere). • Due to their waveguide structure,
there are no bulk optical cavities to be
kept precisely aligned and thermosta-
Figure 3: The ESO LGSF
tized.
projects the laser beam
from behind the M2 hub, • There are no hot temperature spots,
where a 500 mm Launch astigmatisms of optical components,
Telescope is installed. The vibration-induced instabilities.
589 nm laser is mounted • Beam splitters, reflectors and
on the laser clean room dichroics are today all integrated in the
installed under the Nas-
fibre itself giving very high efficiencies.
myth A platform. A single
mode Photonic Crystal • Fibre lasers are very compact and
Fibre relays the laser efficient, and rugged
power for 30 meters, from • They are alignment-free (simplifying
the laser location to the turn-key operation)
focal plane of Launch • They are power scalable up to the
Telescope System behind
power of interest
M2. The LGSF has provi-
sion for 5 such fibres to • They provide in-built fibre delivery
be installed, producing up with diffraction-limited output.
to five Laser Guide Stars. • They have also a potential for low

16 The Messenger 114

 Figure 4: The fibre Raman
laser at 589 nm is obtained
by frequency doubling a laser
at 1178 nm via a second har-
monic generation crystal. The
1178 nm wavelength is
obtained as the first Stokes
of a Stimulated Raman Scat-
tering process. The two con-
figurations of amplifier (upper
drawing) and resonator (lower
drawing) are being bread-
boarded.

cost, whereby current 10 W CW Raman FIBRE RAMAN LASER incident pump photons experience a fre-
lasers cost ~40 kEUR. The Raman laser approach is the sim- quency shift to exactly 1178 nm or twice
plest, more robust fibre laser configura- the sodium D2 wavelength. The efficiency
Moreover, the fibre laser single mode tion and it is directly scalable up to 20 W of this process, on the first Raman Stokes,
output ensures diffraction limited beam CW. It uses the inelastic scattering of pho- is known to reach 80−90% levels in single
qualities, with M2 values ≅ 1.1 (confirmed tons by the molecules of the fibre. The mode fibres. YDFLs are commercially
at high power in our laboratory). The input or pump laser photons excite the available at very high output powers (up
fibre output allows the laser beam to be molecules to a higher vibration state and to 100 W CW), which make them the
relayed around the telescope with diffrac- are thus downshifted in frequency by an ideal pump source for our application. In
tion limited beam quality deployed amount equal to the energy difference a final frequency conversion step (see
directly in the Launch Telescope Systems between the final and initial state of the Figure 4) the 1178 nm output of the
area. This is in most LGSF systems a molecule. This so-called Stokes shift of Raman fibre is frequency-doubled in a
major engineering effort, coping with course depends on the material composi- nonlinear crystal to produce light at 589
laser beam degradation and fast jitter. tion and is, e.g., 17 THz in silica. nm (the second harmonic). Ideally the
With fibre lasers this effort is removed In our fibre Raman laser a ger- frequency doubling happens in a single-
from the LGS telescope facility cost and manosilicate single-mode fibre is used as pass through a periodically poled crystal
complexity. the Raman scattering medium (see Figure (see below). Alternatively, second har-
4). The fibre is pumped by a 40 W CW monic generation by a bulk crystal in a
WHICH FIBRE LASERS commercial ytterbium-doped fibre laser small resonant cavity can be used, a con-
Since there are no optical materials (YDFL) operating at 1121 nm. By this the cept that is used in commercial frequen-
known that are directly lasing at 589 nm,
nonlinear optical frequency conversion
processes have to be exploited in the fibre
Figure 5: Possible output wavelengths out
laser design. of the fibre Raman laser using different
Our approach (Hackenberg et al, pump diodes available on the market as of
2003) is optical frequency-shifting of today. This picture shows the potentiality of a
infrared light by stimulated Raman scat- hydrogen-loaded fibre Raman laser of the
type patented by ESO.
tering (SRS) to exactly twice the sodium
wavelength (or 1178 nm) in combination
with subsequent frequency doubling. This
is a Raman fibre laser which we have
designed and is being breadboarded at
the ESO laser laboratory.
We are also collaborating with LLNL,
which has proposed a sum-frequency
mixing (SFM) from two suitable fibre
lasers at infrared wavelengths, 1583 nm
and 938 nm. Common to both approaches
is the use of periodically poled crystals for
the nonlinear optical frequency conver-
sion of infrared light to the wavelength of
589 nm. Both fibre laser concepts will be
described in more detail in the following.

Bonaccini D., Hackenberg W., & Taylor L., ESO Fibre Laser Development © ESO - December 2003 17
 therefore the potential for many FIBRE LASER
applications besides astronomi- DEVELOPMENT STATUS
cal LGS-AO. This is shown in Raman Laser
Fig. 5, for hydrogen loaded We have done at ESO, and in parallel
fibres. It is likely the most attrac- contracted out to Directed Energy Solu-
tive configuration for industry, as tions, full numerical simulations for the
high power narrow-band Raman Amplifier and Resonator Raman laser
fibre laser do not exist on the configurations. Both indicate feasibility of
market yet. ESO is patenting the the laser system and sufficient suppres-
invention and is applying for sion of the SBS.
technology transfer funds We have procured all discrete compo-
together with interested indus- nents, and assembled the amplifier con-
tries. Our goal is to have compact figuration with discrete components.
and turn-key commercially avail- Thanks to the work of Daniela Werner, a
Figure 6: Numerical Simulations results for the fibre
able fibre lasers for LGS-AO student who has done her master thesis
Raman laser. The 1178 nm output power obtained vs.
pump power is shown, for different output couplers. A
within three years. with us, we have obtained practical expe-
10% output coupler will give the best performance of In Figure 6 the predicted nar- rience with the fibre amplifier and
the fibre laser. 1121 nm pump powers up to 100W are rowband (0.5 GHz) output Raman emissions up to 4 W CW.
now commercially available. power at 1178 nm is shown as a We are now procuring and assembling
function of pump power. At each the all-fibre configurations from industry,
pump power the optimal fibre which will give a better Raman conver-
cy-doubled laser. length maximising the output power has sion efficiency. We have prepared the test
The fine tuning of the output wave- been assumed. This length is less than set-up and the doubling crystal to first get
length and bandwidth to exactly the sodi- 100 m at the highest pump powers shown ~1 W CW of yellow light, and will then
um D2 line wavelength can be obtained in here. With commercially available YDFL scale it up. We have sufficient pump
two ways (see Figure 4), in an amplifier or as pump sources output powers in excess power and all components are specified
a resonator configuration. of 40 W at 1178 nm are feasible. This sets to perform 10 W CW tests in our labs.
an upper limit to the conversion efficien- The breadboard tests are mandatory
Raman amplifier configuration cy needed in the subsequent second har- for the technology transfer to have the
The Raman fibre is seeded with light from monic generation. fibre laser engineered by industry.
a tunable narrowband laser operating at
exactly twice the sodium D2-line wave- SUM-FREQUENCY Sum Frequency
length. The output is the seed light ampli- FIBRE LASER The sum frequency laser has produced
fied by the SRS process in a single pass. The sum-frequency fibre laser is devel- the first yellow light (50 mW) from fibre
This possibility has now become very oped at LLNL, with collaboration from laser obtained in November 2002, and it is
attractive due to the recent commercial ESO. It is based on two rare-earth doped close to the 1 W level today.
introduction of low-power tunable and fibre amplifiers (Fig.7). The 938 nm laser (Dawson et al.,
narrowband 1178 nm laser. Parasitic non- An erbium-doped fibre amplifier 2003) is the technological challenge. It is
linear optical effects like stimulated Bril- (EDFA) operates at 1583 nm, and a getting close to specifications. Problems
louin scattering (SBS) are suppressed by neodymium-doped fibre amplifier to be resolved are the amplified mode
the use of a sufficiently short Raman (NDFA) works at 938 nm. Both fibre selection with the large core NDFA fibre,
fibre. amplifier outputs are mixed in a nonlin- and polarization control. Currently we
ear crystal to generate light at the sum have achieved 4.5 W CW useful output at
Raman resonator configuration frequency corresponding to 589 nm. The 938 nm. We have achieved 5 W CW at
A narrow-band resonator is created with- fibre amplifiers are double-clad pumped 938 nm and 6 W CW at 1583 nm, polar-
in a single mode fibre, using a pair of ded- by high-power diode laser. The seed lasers ized, which at the time of this writing are
icated fibre Bragg gratings tuned to needed for output frequency control are being combined into the sum-frequency
1178 nm wavelength. The Bragg gratings low-power tunable and narrowband crystal.
for the Raman-shifted light are designed diode laser.
such that they only reflect 1178 nm and The EDFA is
not wavelengths that might be created by constructed entire-
parasitic nonlinear processes such as ly of commercially
Stimulated Brillouin Scattering. Thus, available compo-
only the 1178 nm (Raman Stokes I) light nents. The NDFA
will be enhanced in the laser cavity by requires develop-
multiple pass reflections from the Bragg ment to bring it to
gratings, whereas any parasitic nonlinear realisation. This
light will leave the resonator in a single will be discussed in
pass without additional reflections, thus the next Section.
preventing the stimulated SBS emission. ESO contribu-
Again, having the fibre made sufficiently tions are for design
short, SBS will not build up in a single issues on the
pass. 938 nm arm, some
This configuration allows building basic components
narrow band Fibre Raman Lasers at and joint work on
Figure 7: Scheme of the sum-frequency fibre laser pursued at LLNL.
many other wavelengths than 589 nm, and the non-linear The sum of 1583 nm and 938 nm photons energies creates photons
also multiple wavelength lasers. It has crystals. at 589 nm in a Periodically Poled SFG crystal.

18 The Messenger 114

 938 nm Neodimyum
Doped Fibre Amplifier
In the NDFA design the main technical
challenge is the suppression of amplified
spontaneous emission (ASE).To force the
amplifier to operate at 938 nm and to sup-
press the ASE at longer wavelengths we
are using in parallel several methods:
seeding at high power, introducing bend
losses to the longer wavelengths which is
especially effective in fibres of low
numerical apertures and finally long-peri-
od Bragg gratings for frequency-selective
feedback in the amplifier fibre. These
Figure 8: Alex at LLNL tuning up the laser to get the maximum output at the time methods have different efficiencies, but
(Nov 6, 2002). This is the first yellow light from a fibre laser, produced at 589 nm more or less independent from each
with 1 GHz bandwidth, 50 mW. other. We have designed and developed a
special fibre to use for the 938 nm laser,
TECHNOLOGIES RELATED The periodically poled crystals com- together with the Institut für Physikalis-
In pursuing the fibre laser we are pushing mercially available are sensitive to the so- che Hoch Technologie in Jena (Germany)
some technologies which are related to it, called photorefractive damage. Together
together with industry. These are with LLNL we are investigating several WHAT ARE WE GOING TO DO NEXT
described in the following. methods to avoid this type of damage. ON THE FIBRE RAMAN LASER
This include elliptical beam formatting to For the ESO Raman fibre laser we will
Narrow band high power keep the power density low and improved receive the all-fibre lasers breadboard
Raman lasers (i.e., stochiometric) crystal growth and around mid 2004, and perform the tests in
Although high power Raman lasers are poling techniques. For this purpose a crys- the summer in our laboratory. The period-
commercially available at 1178 nm with tal test facility has been set up at LLNL to ically poled crystals activities shared at
powers up to 15 W CW, their linewidth is investigate conversion efficiency and LLNL will help us to choose the appro-
of the order of 1 nm. We are pushing the long-term reliability. European laser priate crystal for the Second Harmonic
linewidth to 1 pm, the main challenge in industries have expressed strong interest Generation required by the Raman laser.
doing this is to suppress the Stimulated in this technology. We are investigating We will in the meantime apply for
Brillouin Scattering. For this we have a PPSLN and PPSLT materials. European funds for technology transfer
viable path supported by design, numeri- of these new laser types to industry. We
cal simulations and soon by experiments. Fibre Bragg Gratings are forming a consortium of companies
Fibre Bragg Gratings (FBG) are reflec- with member state countries and prepar-
High power fibre couplers tive optical elements or filters created by ing the funding application for the begin-
Beam splitters and dichroics are fully periodic modulation of the refractive ning of 2004. The deliverable will be fibre
integrated in the fibre laser, in the form of index in the fibre core. They are written lasers at 589 nm, to be field tested at our
TAP couplers and WDM. Recently high permanently into the glass of the fibre by LGSF facility. An option for pulsed for-
power (20 W CW) couplers have been exposing it once to a suitable ultraviolet mats fibre lasers will be opened.
developed and are commercially avail- light interference pattern. The first part of the program will last
able. Couplers and dichroics for polarisa- Apodized Fibre Bragg gratings of the two years leading to a tested, well engi-
tion preserving fibres are still under required extinction and shape for the neered preproduction unit. The second
development by the industry. Although Raman fibre laser are feasible, although part will call for a field test final unit, to
not mandatory, they will be an asset for our specifications have been considered be delivered in the subsequent year.
the fibre Raman laser. by industry as high-end, low-yield FBGs.
REFERENCES
Periodically-poled Beuzit, J-L., Demailly, L., Gendron, E. et al.,
1997, in Experimental Astronomy, 7, 285
nonlinear crystal Beckers, J.M., 1998, in Proc. “ESO Conference
Periodically-poled crystals have on Very Large Telescopes and Their Instru-
the advantage of a very high nonlin- mentation”, ed M-H Ulrich, p. 693
ear coefficient. They can be inserted Dawson J.W., Drobshoff A., Liao Z.M. et al.,
2003, in ”Advances in Fibre Lasers” ed.
directly in the laser beam and give L.N. Durvasula, SPIE Proc. 4974, p. 75
efficient conversions. Such a crystal Ellerbroek B., Rigaut F., Bauman B. Et al.,
consists of a sequence of permanent 2003, in ”Adaptive Optical System Tech-
ferroelectric domains of alternating nologies”, eds P. Wizinowich and D. Bonac-
polarity perpendicular to the optical cini, SPIE Proc. vol 4839, p. 55
Gavel D., Gates E., Max C. et al., 2003, in
axis (see Figure 9). By this the sign of ”Adaptive Optical System Technologies”,
the nonlinear coefficient is reversed op. cit., p. 354
every period, and quasi-phase Hackenberg W., Bonaccini D. and Werner D.,
matching is achieved for the three 2003, in ”Adaptive Optical System Tech-
nologies”, op. cit., p. 421
interacting waves in the crystal. Takami H., Takato N., Hayano Y. et al., 2003, in
Without poling the fundamental and ”Adaptive Optical System Technologies”,
second harmonic waves would Figure 9: Periodically poled cristals allow to op. cit. p. 21
quickly run out of phase due to chro- obtain sum frequency or frequency doubling Wizinowich P.L., Le Mignant D., Stomski P.J.Jr.
matic dispersion. effects, without the need of a resonant cavity. et al., 2003, in ”Adaptive Optical System
They are an emerging laser technology. Technologies”, op. cit., p. 9

Bonaccini D., Hackenberg W., & Taylor L., ESO Fibre Laser Development © ESO - December 2003 19
 S ETTING N EW S TAND ARDS
WITH HARPS
BY OCTOBER 1ST, 2003, ESO'S NEW AND UNIQUE PLANET-HUNTING MACHINE HARPS (HIGH-ACCURACY
RADIAL VELOCITY PLANETARY SEARCHER) HAS BECOME OPERATIONAL. THE MEASUREMENTS MADE DURING THE
COMMISSIONING PHASE AND THE FIRST WEEKS OF OPERATION ARE OF OUTSTANDING QUALITY. IN THIS ARTICLE WE
REPORT AMONG OTHER EXAMPLES ON THE FIRST EXTRA-SOLAR PLANET DISCOVERED WITH HARPS AND ON THE
DETECTION OF TINY STELLAR OSCILLATIONS. THE RESULTS PRESENTED DEMONSTRATE THAT HARPS IS CURRENT-
LY THE MOST PRECISE DOPPLER-MEASUREMENTS MACHINE IN THE WORLD. WITH THIS ACQUISITION ESO PLACES
ITSELF AT THE HEAD OF A SCIENTIFIC DOMAIN, WHOSE INTEREST HAS CONTINUED TO GROW DURING THE PAST
YEARS.

HARPS PROJECT was AN UNEQUALLED FACILITY FOR

T
HE
M. M AYOR 1, F. P EPE 1, born in May 1998 when ESO RADIAL-VELOCITY MEASUREMENTS
D. Q UELOZ 1, F. B OUCHY 2, issued an Announcement of HARPS is a fibre-fed, cross-dispersed
Opportunity asking for the echelle spectrograph. Two fibres, an object
G. R UPPRECHT 3, G. L O design, construction, and pro- and a reference fibre, feed the spectro-
C UR TO 4, G. AVIL A 3, curement of an instrument dedicated to graph with the light from the telescope and
the search for extrasolar planets and aim- the calibration lamps. The fibres are re-
W. B ENZ 5, J.-L. B ER TAUX 6,
ing at an unequalled precision of 1 m/s. In imaged by the spectrograph optics onto a
X. B ONFILS 1, T H . DALL 4, response to ESO’s call the Observatoire de mosaic of two 2k4 CCDs, where two
H. D EKKER 3, B. D EL ABRE 3, Genève together with the Physikalisches echelle spectra of 72 orders are formed for
Institut der Universität Bern, the Observa- each fibre. The covered spectral domain
W. E CKER T 4, M. F LEURY 1, toire de Haute-Provence, and the Service ranges from 380 nm to 690 nm. The resolu-
A. G ILLIOTTE 4, D. G OJAK 3, d’Aéronomie du CNRS have formed a tion of the spectrograph is given by the
Consortium which realized, in collabora- fibre diameter and attains a value of about
J.C. G UZMAN 4, D. KOHLER 7, tion with ESO, this ambitious project in R=115,000. At this resolution each spectral
J.-L. L IZON 3, A. L ONGINOTTI 3, only three and a half years. The instrument element is still sampled by 3.2 CCD pixels.
was installed by the Consortium on ESO’s A summary of the main HARPS features
C. L OVIS 1, D. M ÉGEVAND 1, 3.6-m Telescope at La Silla in January 2003 is given in Table 1. For a detailed descrip-
L. PASQUINI 3, J. R EYES 3, and first light took place on February 11th. tion of the instrument we refer to the arti-
HARPS was commissioned during the cle published in Pepe et al. (2002) and to
J.-P. S IVAN 7, D. S OSNOWSKA 1, periods 70 and 71. During period 71 a first the HARPS Users Manual.
R. S OTO 4, S. U DRY 1, GTO run of seven nights had been allocat- HARPS is an ordinary spectrograph
ed for the Consortium as well. These with outstanding efficiency and spectral
A. VAN K ESTEREN 3,
observations have already produced many resolution. The main characteristics of
L. W EBER 1, U. W EILENMANN 4 exciting results. HARPS is however its extraordinary sta-
The instrument was handed over to bility. The ThAr-reference technique is
ESO La Silla by the end of September able to measure and correct the tiniest
1OBSERVATOIRE DE 2003 and made available to the Communi- instrumental drifts; nevertheless a lot of
GENÈVE, ty for period 72. Together with the 3.6-m effort was put into making the spectro-
SWITZERLAND; telescope HARPS is now delivering fan- graph intrinsically stable, in order to avoid
2L ABORATOIRE D'ASTROPHYSIQUE tastic scientific frames to
DE MARSEILLE, FRANCE; the observers. But this is
Table 1: HARPS spectrograph characteristics
3ESO GARCHING; not the only data product
of HARPS. The powerful
4ESO L A SILLA; Optical design fibre-fed, cross-dispersed
HARPS pipeline provides echelle spectrograph
5PHYSIKALISCHES INSTITUT DER any observer with extract-
Technique simultaneous ThAr Reference
UNIVERSITÄT BERN, SWITZERLAND; ed and wavelength cali-
Number of fibres 2
6SERVICE D'AÉRONOMIE, brated high-resolution
spectra, as well as the pre- Fibre aperture on sky 1 arcsec
VERRIÈRES LE BUISSON, FRANCE; cise radial velocity of the Collimated beam diameter 208 mm
7OBSERVATOIRE DE HAUTE- observed star. The observ- Covered spectral range 380 nm to 690 nm
PROVENCE, FRANCE er will thus leave the Spectral resolution R=115,000
Observatory with fully
Spectral format 72 echelle orders
reduced spectra. Addi- 61.44 x 62.74 mm
tional information on how CCD chip mosaic, 2xEEV2k4
to observe with HARPS pixel size=15µm
can be found at Sampling 3.2 pixels/SE
http://www.ls.eso.org/lasilla/ Min. inter-order 33 pixels
sciops/harps/.

20 The Messenger 114

 The extraordinary stability translates
directly into the stability of the radial
velocity measurements. This is most
impressively shown by the measurement
series presented in Figure 2b. Both fibres
have been exposed repetitively with the
ThAr calibration light and the drift –
expressed in m/s – was computed as a func-
tion of time. During several hours the total
drift remains well below 1 m/s and the
measurement is dominated by “noise”. This
noise is introduced by the CCD whose
temperature varies by ±0.02 K and pro-
duces microscopic dilatation of the chip. In
fact, both fibres show the same behaviour
and when subtracting one fibre from the
other this “noise” disappears and we
obtain the dispersion values expected from
the photon-noise level of the simultaneous
ThAr reference. These measurements
prove that the simultaneous ThAr refer-
ence technique is perfectly able to track
instrumental drift at a level of 0.1 m/s rms.
Figure 1:The HARPS spectrograph inside the air-conditioned room at La Silla just before being HARPS has been optimised for radial-
closed and evacuated. velocity efficiency, i.e. to obtain in the
shortest possible exposure time the most
as far as possible any kind of second-order uum vessel controls the air temperature to precise radial-velocity measurement. How-
instrumental errors. As a consequence, we 17!C with a long-term stability of the order ever, this requirement is not necessarily
have decided to operate the spectrograph of 0.01!C. Because of the huge thermal synonymous with optical efficiency. For
in vacuum (see Figure 1), since ambient inertia of the vacuum vessel and the excel- example, spectral resolution is an impor-
pressure variations would have produced lent thermal insulation between spectro- tant factor in reducing the photon noise of
huge drifts (typically 100 m/s per mbar). graph and vessel provided by the vacuum, the radial-velocity measurement but is
The operating pressure is always kept the short-term temperature stability obtai- often in competition with the optical effi-
below 0.01 mbar such that the drifts never ned is even better. Over one day we have ciency of the spectrograph or with the size
will exceed the equivalent of 1 m/s per day. measured variations of the order of 0.001 K of the instrument and the telescope. On
Not only the pressure but also the tem- rms. An example of the excellent control is the other hand the instrumental errors
perature is strictly controlled. In fact, a given by the behaviour of the echelle grat- must also be low, at least as low as the goal
two-stage air-conditioning around the vac- ing temperature plotted in Figure 2a. for the photon-noise equivalent. The

Figure 2: a) Tempera-
ture of the echelle
grating over 24 hours
recorded at two differ-
ent days. b) Drift of
the spectrograph dur-
ing a ThAr series of
several hours proving
that the spectrograph
is extremely stable
and that the simulta-
neous ThAr reference
is able to track and
correct tiniest instru-
mental drifts.

Figure 3: a) Total effi-

ciency of HARPS
(right-hand ordinate)
and efficiency of single
components (left-hand
ordinate). b) Compari-
son of the measured
SNR with the values
calculated by means of
the Exposure Time
Calculator. For com-
parison we performed
a 71s exposure on
HD10700 (mv=3.5) at
an airmass of 1.1. The
DIMM monitor was
indicating a seeing of
0.9 arcsec.

© ESO - December 2003 21

 1.7 m/s and partly due to photon noise.
The residual noise is a combination of cal-
ibration errors (< 0.5 m/s) and effects due
to the star itself, which can be due to pul-
sations, activity and jitter. It is also worth
noting that the HARPS vaccum has been
broken between the two commissioning
runs.
We conclude that the short-term pre-
cision (1 night) of HARPS is well below 1
m/s. Even on the time scale of a commis-
Figure 4: a) Series of 7 hours and 420 exposures on α Centauri B proving the extraordinary
short-term precision of HARPS. b) Zoom of figure a) to illustrate the presence of a periodic sig-
sioning period (2 weeks) we have been
nal produced by the stellar pulsation. able to measure standard stars which
showed radial-velocity dispersions of
1 m/s including photon noise. The same
level of precision is obtained on
HD 83443 over a time scale of 4 months.
Apart from photon noise a major contri-
bution to the residuals probably comes
from the star itself (pulsations, activity, jit-
ter). Only an in-depth, long-term study
will allow us to identify the relative
importance of the various error sources.

Figure 5: a) Power spectrum of α Cen B. The acoustic modes corresponding to the 4-minutes
CATCHING THE TINY MELODY
oscillation are clearly identified and emerge well above the noise. b) Autocorrelation of the OF A SOLAR-LIKE STAR
power spectrum of α Centauri B. Stars, which are spheres of hot gas, prop-
agate very well in their interiors acoustic
design of HARPS is consequently the that the average contribution of modes to waves which are generated by turbulent
result of an accurate trade-off between all the total dispersion reaches 0.44 m/s. By convection near their surfaces. Frequen-
the relevant factors. Despite various com- quadratic subtraction of this value from cies and amplitudes of these acoustic
promises regarding for example the fibre the measured dispersion we obtain a waves, also called oscillation modes or p-
diameter (1 arcsec on the sky) and the noise level of 0.26 m/s, which is in good modes, depend on the physical conditions
fibre-feed itself, the optical efficiency agreement with the value extrapolated prevailing in the layers crossed by the
obtained is remarkable. Figure 3a shows from the mean white noise level meas- waves and provide a powerful seismolog-
the calculated total optical efficiency of ured in the high frequency range between ical tool. Helioseismology, which moni-
the instrument and some of its subsys- 6 and 8 mHz in the power spectrum. The tors the oscillation modes of our Sun, has
tems. These values have been adopted in photon noise on a single measurement is been used since the 1970’s and led to
the current version of the HARPS-Expo- of 0.17 m/s, which leaves less than 0.2 m/s major revisions in the “standard model”
sure Time Calculator (ETC) which serves for all other possible error sources (ThAr of the Sun and provided, for instance,
to estimate the SNR obtained during an noise, guiding errors, influence of the measures of the Sun’s inner rotation, the
exposure. In Figure 3b we compare the atmosphere, instrumental errors). size of the convective zone and the struc-
results obtained using the ETC with real The long-term precision of the instru- ture and composition of the external lay-
measurements on the star HD 10700. ment cannot be checked easily because it ers. Solar-like oscillation modes generate
requires a long time base on one hand, periodic motions of the stellar surface
HARPS DELIVERING PRECISION and the knowledge of stable stellar with periods in the range 3 − 30 min., but
NEVER REACHED BEFORE sources on the other hand. Especially the with extremely small amplitudes. The cor-
During the HARPS commissioning we latter point represents a new challenge responding amplitudes of the stellar sur-
monitored the star α Centauri B, which is since the intrinsic stability of the stars has face velocity modulations are in the range
a K-type star substantially smaller than never been studied at this level of preci- 10 − 100 cm/s.
our Sun. During 7 hours we collected a sion before. Nevertheless we have
total of 420 spectra with typical SNR of been able to gain some indication of
500 each at λ=550 nm. The radial-velocity the long-term precision of HARPS
measurement sequence plotted in Figure by observing the star HD 83443
4a indicates a dispersion of 51 cm/s but which has a planetary companion
the zoom shown in Figure 4b shows that with well known orbital parameters.
this dispersion is completely dominated Figure 6 shows a total of nine radial-
by 4-minutes stellar oscillations. In fact, velocity data points, some of them
the power spectrum of this sequence collected during the first commis-
shown in Figure 5 clearly exhibits a series sioning period in February 2003, and
of peaks around 4 mHz, corresponding to some others about 4 months later
individual acoustic modes of the star, with during the second commissioning
amplitude in the range 10-20 cm/s. The held in June 2003. All the data fit
positive interference of several oscillation well the calculated orbit whose
Figure 6: Radial-velocity data of HD 83443,
modes may lead to amplitudes much larg- parameters are fully consistent with
which harbours a known extra-solar planet. The
er than the amplitude of single modes. the previously known solution. The data fit well the calculated orbit with a weighted
From the power spectrum we estimate obtained dispersion (O−C) is only dispersion of 1.7 m/s.

22 The Messenger 114

 Two years ago, we were able to detect Table 2: Stellar parameters and oscillation modes properties
with the CORALIE spectrograph an
Spectral Peak to Peak
unambiguous oscillation signal of 31 cm/s Star Period RMS Dispersion
type modulation
amplitude of the star α Centauri A, a
nearby solar twin (Bouchy & Carrier, α Cen B K1V 4 min. 1.5 m/s 0.5 m/s
2001 & 2002). With these measurements
several oscillation modes were separated HD20794 G8V 5 min. 2.5 m/s 0.8 m/s
and clearly identified.
The measurements made on α Cen-
HD160691 G5IV-V 9 min. 5 m/s 1.7 m/s
tauri B with HARPS have lower frequen-
cy resolution because of the limited dura-
β Hydri G2IV 16 min. 5 m/s 1.7 m/s
tion of the tests. To obtain the same reso-
lution we would have had to observe the
star for 13 consecutive nights. On the
other hand the measurement obtained sequences (several nights are needed to interference leads to modulation of the
with HARPS are impressive mostly reach a sufficient frequency resolution to Doppler time series of up to 10 times
because of efficiency, cycle time and pre- characterize individual p-modes) of high their value. This is clearly a limitation for
cision. Figure 4 illustrates clearly the sampling radial-velocity measurements high precision exoplanet surveys at the
detectivity obtained with HARPS: will permit to measure acoustic waves level of 1 m/s and requires us to define a
despite the small amplitude the 4-minutes with amplitudes as tiny as a few cm/s. precise strategy of target selection and
stellar oscillation can be detected “by observation. Short-period oscillations (4−
eye” directly in the time series, and they DOWN TO THE STELLAR LIMIT 8 min.) of small amplitudes, which occur
become even more evident in the Fourier After the amazing results obtained on α in G and K dwarfs, can be easily averaged
Transform space shown in Figure 5a. Fig- Centauri B we have decided to monitor a out with typical exposures times of 15
ure 5b shows the autocorrelation of the small set of solar type stars. Figure 7 minutes. Sub-giants and giants, which can
power spectrum. The plot clearly indi- shows short sequences of radial velocity show larger amplitudes and longer peri-
cates that the peaks have a comb-like measurements obtained on these objects. ods, should however be carefully
structure, with a main peak separation of On each of these sequences the stellar removed from the sample. This kind of
about 160 µHz. This is the typical signa- oscillations are clearly visible. Stellar precaution should permit us to detect
ture of oscillation modes of identical parameters and observed oscillation exoplanets at the level of 1 m/s.
angular degree expected for solar-like modes are listed in the Table 2. As expect-
stars. The intermediate peaks appearing ed, these measurements clearly show that OBSERVING FAINT OBJECTS
in the figure correspond to the correlation period and amplitude of the oscillation The OGLE-III programme, involved in
between modes with angular degree l=1 modes are directly related to the stellar an extensive ground-based photometric
and l=0 or 2. properties. They demonstrate the full survey for planetary and low luminosity
Such a level of accuracy in the detec- capability of HARPS for asteroseismolo- object transits, has recently announced
tion of solar-like oscillation modes has gy, not only on very bright stars. the detection of transiting candidates
never been reached before. This result These results show however also that around stars located in the direction of
obtained with HARPS on a very short the main limitation for HARPS will come the Galactic centre. The radii of these
sequence (7 hours), clearly shows the from the stellar “noise” itself. The individ- objects were estimated to be in the range
unique potential of this instrument in the ual mode amplitudes of these stars are in between 0.5 and 5 Jupiter radii, meaning
domain of asteroseismology. Long the range 10−50 cm/s but their additive that some of these objects are probably in
the planetary domain. In order to meas-
ure the mass of these objects, and then
determine their real nature, a radial
velocity follow-up is necessary.
In order to test the limiting magnitude
of HARPS, we observed the faint candi-
date OGLE-TR-56 (mv=16.6). One hour
exposure allowed us to reach a signal-to-
noise ratio between 3 and 4 correspon-
ding to a photon noise uncertainty on the
radial velocity of about 30 m/s. The five
RV measurements presented in Figure 8
demonstrate the full capability of
HARPS to find and characterize hot
Jupiters around stars as faint as OGLE-
Figure 7: TR-56. We have fitted to these data points
Summary plot a sinusoidal curve with fixed period as
of the radial- given by the photometric study. The meas-
velocity of ured amplitude is however much different
stars for which
we have made and in contrast to the results published by
long observa- Konacki et al. (2003). A complete analysis
tions series. of these measurements will be required in
All of them order to determine whether this RV mod-
show clear
indications of
ulation is due to a planetary companion
star pulsation. or to a binary system blended and

Mayor M. et al., Setting New Standards with HARPS © ESO - December 2003 23
 velocity measurements of several hun- ed to the mass of the convective zone and
dred stars. Many of them show a varying probably finds its origin in the chemical
radial velocity, sometimes with a clear composition of the primordial molecular
periodicity. In the case of HD 330075 the cloud. To add new constraints to the link
signal is most probably due to an exo- between star chemical composition and
planet orbiting this star in 3.37 days. The frequency (or properties) of exoplanets,
radial-velocity curve of HD 330075 is we will carry out two additional pro-
shown in Figure 9. Precise photometric grammes: The first programme is a search
measurements carried out on the SAT for exoplanets orbiting solar-type stars
Danish Telescope and by the Swiss-Euler with notable deficiency (for most of them
Telescope have shown the absence of a [Fe/H] between −0.5 and −1.0). Among
photometric transit, which would have the existing detections of exoplanets only
allowed us to determine the radius and two or three have been found with metal-
the orbital inclination of the planet. Nev- licity in that range. We aim at being able
ertheless, these measurements have also of estimating the frequency of exoplanets
Figure 8: Radial-velocity of OGLE-TR-56
demonstrated that the photometric vari- in that domain of metallicity and, if possi-
measured with HARPS. Despite the faint mag-
nitude of the object, the dispersion of the data ability is low, and that we can exclude ble, to compare their characteristics
points around the fitted curve is only 30 m/s. therefore the radial-velocity variation to (masses, orbits) to planets orbiting metal
be produced by stellar activity. This makes rich stars. The second programme aims at
HD 330075 the first HARPS extra-solar exploring the link between stellar metal-
masked by the main star. This kind of planet candidate. licity and properties of exoplanets. Visual
study has a huge interest for the future Only a few of the hundred detected binaries with solar-type stars of almost
radial-velocity follow-up of exoplanets planets have masses less than the mass of identical magnitudes have been selected.
detected by the satellite COROT. Saturn, and due to the present precision We will search for exoplanets orbiting
of radial velocity surveys the distribution one of the components of these systems.
THE HARPS of planetary masses is heavily biased (or For those including giant planets a
EXOPLANET PROGRAMME completely unknown) for masses less detailed chemical analysis will be done
The HARPS Consortium GTO pro- than half the mass of Jupiter. We will take for both stellar components to search for
gramme will be devoted exclusively to the advantage of the very high precision of possible differences in their chemical
study and characterization of exoplanets, HARPS to search for very low mass plan- compositions.
in continuation of a planet-search pro- ets. For a sample of pre-selected non- Follow-up radial velocity measure-
gramme initiated 10 years ago (Queloz & active solar-type stars we will be able to ments for stars with planetary transits
Mayor, 2001). For a large, volume-limited explore the domain of the mass-function detected by the COROT space mission
sample we will do a first screening in for planetary masses less than the mass of will be made with HARPS. Photometric
order to identify new “Hot Jupiters” and Saturn down to a few Earth masses for transits provide an estimate of the radius
other Jovian-type planets. Increasing the short periods. of the transiting planet as well as the
list of “Hot Jupiters” will offer a chance to A systematic search for planets will be orbital period and phase. Complementary
find a second star with a planetary transit made for a volume-limited sample of M- ground-based spectroscopic measure-
among relatively bright stars. Better sta- dwarfs closer than 11 parsecs. Such a sur- ments with HARPS will constrain the
tistics are needed to search for new prop- vey of very low mass stars will give us a planetary mass and then the planet mean
erties of the distribution of exoplanet chance to derive the frequency of planets density. The main scientific return for the
parameters. as a function of the stellar mass. These planetary programme of the COROT
This part of the programme already objects are of prime importance for mission will come from the combination
started in July 2003 within the first GTO future astrometric studies to be carried of the photometric and radial velocity
period assigned to the Consortium. In this out with the VLTI or space-mission like data.
short run of only 9 nights HARPS SIM.
unveiled all its characteristics, its optical Stars with detected giant planets ACKNOWLEDGEMENTS
We would like to acknowledge the many peo-
efficiency, the unique precision and, in exhibit an impressive excess of metallicity ple whose names do not appear among the
particular, the outstanding efficiency of in contrast to stellar samples without authors but have contributed to the great suc-
the reduction pipeline. In fact, we have giant planets (Santos et al., 2001). The cess of the HARPS project through their valu-
collected about 500 spectra and radial excess of metallicity does not seem relat- able and dedicated work. This project has been
financed by the Swiss National Science Foun-
dation, the French region “Provence, Alpes et
Côte d’Azur”, the Institut National des Sciences
de l’Univers INSU (F), the Université de
Figure 9: First HARPS Genève (CH), the Observatoire de Haute-
extra-solar planet candidate Provence (F), the Physikalisches Institut der
HD 330075. The planet has a Universität Bern (CH), and the Service
minimum mass of 0.8 MJ and d’Aéronomie du CNRS (F). The HARPS Pro-
orbits the star in 3.37 days at ject is a collaboration between the HARPS
a distance of 0.044 AU from Consortium and the European Southern
its parent star, and belongs Observatory (ESO).
therefore to the category of
Hot Jupiters. The induced REFERENCES
radial-velocity variation has a Bouchy F. & Carrier F., 2001, ESO Messenger,
semi-amplitude of 107 m/s. 106, 32
The total dispersion with Bouchy F. & Carrier F., 2002, A&A, 390, 205
regard to the orbital solution Konacki M. et al., 2003, Nature, 421, 507
is (O-C)=2.5 m/s, and includes Pepe F. et al., 2002, ESO Messenger, 110, 9
photon noise, stellar pulsa- Queloz D. & Mayor M., 2001, ESO Messenger,
tion and possible jitter. 105, 1
Santos N. et al., 2001, A&A, 373, 1019

24 The Messenger 114

 LISA GERMANY
SciOps

DOCUMENTATION of this, as far as the community is con- SEST

Those of you who applied for WFI time in cerned, is that observers will now use In the June Messenger we brought you
P73 will (hopefully!) have noticed the the ESO standard P2PP to observe with highlights from the last dishwalk at the
new-look WFI web pages (you do read FEROS. This will also allow FEROS to be SEST. Since the end of August however,
them don't you?). These pages should operated in Service Mode, opening the the SEST has been stowed and we have
be easier to navigate, and you should be door for new types of projects (e.g. long had to bid the staff of SEST farewell from
able to find the information you are after term monitoring). This will result in a sig- La Silla. Some of them can now be
much quicker than in the past. nificant increase in productivity of the found moving further up in the world (in
A similar thing has just happened for 2.20m telescope since the constraints both latitude and altitude) and we wish
SUSI2 on the NTT. The SUSI2 webpages for WFI and FEROS programmes are them every success with APEX.
have just had a complete revamp to largely complimentary.
make important information much easier The first stage of this upgrade, the
to find. Fibre Service mission, was carried out in MARS
As a result of two successful October. This involved a complete 27th August 2003 found Mars at the
upgrades, TIMMI2 has a new user's replacement of the Science fibres, and closest it has been to Earth in a very
manual available online. A postscript has successfully restored FEROS to its long time, and the staff of La Silla as well
version will soon be released. intrinsic high efficiency (approximately as the kids from the village of Cachiyuyo
In addition, the postscript version of 18% for Telescope+Instrument+Detec- at the ESO 1m telescope to witness the
the EFOSC2 user manual has had a tor). We now finally realise the factor 2 event. A brief presentation explaining the
major overhaul for the first time since increase in throughput (compared to the astronomy behind the close approach
1999. The new manual can be found ESO-1.52m) due to the increased light preceded the viewing session through a
under the "Documentation" link on the collecting area of the 2.20m. mounted eyepiece (not often you actual-
EFOSC2 webpage. The November commissioning peri- ly get to look through a 1m telescope
One way to improve our documenta- od should see the complete implemen- these days!). The seeing conditions were
tion and make it more useful to the user tation of the VLT standard Instrument really quite good and you could clearly
community is for you to provide us with and Detector Control Systems, and the see lighter and darker regions on the
feedback. If you have any comments you replacement of the current Copenhagen surface of Mars, as well as the bright,
would like to make on any of the La Silla University Astronomical Observatory white polar cap. La Silla University really
documentation, please email us at lasil- CCD Controller with an ESO standard put on a wonderful evening for all who
[email protected] FIERA CCD Controller. could attend, and it was amazing to see
The new FEROS (FEROS-II) will be such a lot of detail on the surface of
back online in mid-November and will Mars with my own eyes. Thanks to our
SOFI
immediately be put to work on several resident photographer, Peter Sinclaire,
The InfraRed spectro-imager at the NTT
Service Mode programmes. The first vis- photographs of the event can be found
has not been at all well over the past few
iting astronomer to use FEROS-II will at:
months, suffering several mishaps one
arrive in late December. http://www.pbase.com/psinclai/marte&page=all
after the other. Several wheels have
either lost their initialization switches or
become stuck, making the operation of
SofI quite challenging. The final days of
October will see SofI undergo a thor-
ough maintenance mission, during
which time the instrument will be almost
completely dismanteled and three of the
wheels replaced. We expect that this
intervention will restore SofI's reliability.

FEROS (John Pritchard)

FEROS is once again undergoing a
major upgrade. After last year's suc- The kids from
cessful move from the ESO-1.52m to the Cachiyuyo and the line
MPG/ESO-2.20m, FEROS, the last non- of La Silla staff all wait-
ing to glimpse a peek
VLT compliant instrument, is now being at Mars
given a "VLT makeover". The major result

© ESO - December 2003 25

Reports from Observers
H AR VES TING S CIENTIFIC R ESULTS
WITH THE VLTI
THE ESO VERY LARGE TELESCOPE INTERFEROMETER (VLTI) HAS BEEN INCLUDED FOR THE FIRST TIME IN THE
OFFICIAL CALL FOR PROPOSALS REQUESTING ESO TELESCOPES FOR THE PERIOD STARTING IN APRIL 2004. THIS
MARKS THE OFFICIAL START OF PUBLIC INTERFEROMETRIC OBSERVATIONS OPEN TO THE COMMUNITY. IT IS THE
START OF A NEW APPROACH TO INTERFEROMETRY AS A STANDARD ASTRONOMICAL TECHNIQUE, AND A POINT OF
PRIDE AND SATISFACTION FOR ALL THE PEOPLE WHO HAVE BEEN WORKING WITH THIS CHALLENGING GOAL FOR
MANY YEARS. BUT IT SHOULD NOT BE FORGOTTEN THAT THE VLTI HAS ALREADY LOGGED OVER TWO YEARS OF
INTENSIVE COMMISSIONING, AS WELL AS SOME INITIAL SCIENCE DEMONSTRATION RUNS. OVER 16,000
OBSERVATIONS OF HUNDREDS OF OBJECTS HAVE BEEN COLLECTED AND ARE AVAILABLE PUBLICLY OVER THE ESO
ARCHIVE ON THE WEB. IN 2003, THE FIRST SCIENTIFIC RESULTS OF THIS REMARKABLE EFFORT HAVE APPEARED.
ALREADY MORE THAN A DOZEN PAPERS BASED ON VLTI DATA HAVE BEEN SUBMITTED OR ACCEPTED BY REFEREED
JOURNALS, WITH A SIMILAR VOLUME OF CONTRIBUTIONS TO WORKSHOPS AND CONFERENCES OF A SCIENTIFIC
NATURE. WE PROVIDE HERE AN OVERVIEW OF THIS EARLY SCIENTIFIC PRODUCTION OF THE VLTI, RANGING FROM
THE DETERMINATION OF FUNDAMENTAL PARAMETERS OF MANY CLASSES OF STARS TO THE FIRST INTERFEROMET-
RIC MEASUREMENT OF THE INNER REGIONS OF THE NUCLEUS OF THE SEYFERT GALAXY NGC 1068.

A NDREA R ICHICHI AND F R ANCESCO PARESCE (ESO)

W
HAT DOES A BOTANICAL GARDEN Table 1. Statistics of the VINCI commissioning observations (up to
share with astronomy? Not much August 2003). A total of 321 independent objects have been observed.
at first glance, granted. But in
Munich you find a wonderful Number Total num- Number Volume
Year
example of the former, and many of OBs ber of files of nights (Gb)
professionals of the latter. Strolling through the centu-
ry-old trees and the colourful flowerbeds of the 2001 4827 19308 206 25.2
Botanischer Garten, one wonders how different
plants display different growth patterns. The bamboo
grows 30 centimeters per day in the tropical jungle, 2002 4966 19864 235 35.9
and at the other extreme the canadian white cedar
(Thuja occidentalis) takes one and half century to 2003 6125 24500 180 56
reach a height of 10 centimeters. Then the thought
strikes us: is it not the same with astronomical
research? Some ideas produce wonderful results Total 15918 63672 621 117.2
almost immediately, while some others have to wait
patiently for decades before becoming accepted. And
of course, some ideas never succeed at all.
If the ESO Very Large Telescope Interferometer
(VLTI) were a plant, it would be indeed a strange mix.
Its first seed was planted about twenty years ago when Euro- has been added, a new branch has been spawned. Even the
pean astronomers started to be fascinated by the idea of creat- astronomers and engineers in the VLTI group have trouble
ing a quadruplet of four identical giant telescopes and combin- sometimes keeping up with the news from one side or the other
ing them interferometrically. A visionary concept that took of the ocean, be it the addition of a new delay line or the inau-
decades of engineering feats to become reality. One by one the guration of a new baseline.
four VLT telescopes were erected, changing forever the skyline As 2003 comes to an end, more than two years of patient
of Cerro Paranal. As the giant mirrors swept the skies, the first commissioning with the test instrument VINCI have been com-
incredible images began to open new opportunities for pleted, accumulating over 16,000 observations of hundreds of
astronomers around the world. stars and making them available publicly to the whole commu-
But underground, in tunnels and isolated rooms, other activ- nity (see Table 1). The small siderostat test telescopes have been
ities continued as ESO astronomers and engineers assembled the workhorses of this enormous effort, but the Unit Telescopes
hundreds of mirrors, fine mechanical mounts, and scores of com- (UTs) have been used as well and are ready to be offered pub-
puters: the skeleton of the most powerful interferometer was licly. The first VLTI scientific instrument, MIDI, has been
growing rapidly. Finally, in March 2001, the VLTI saw its first opened to the community starting from Period 73, while the sec-
fringes. From that moment, the growth has been definitely bam- ond one, AMBER, is completed and awaiting shipment from
boo-like. Every few months, or even every few weeks, a new bud Europe. Schools and workshops have been organized around

26 The Messenger 114

Figure 1: Radii and masses
of the four very-low-mass
tion for many interested readers. Typical
stars now observed with the angular size scales of the closest stars and
VLTI, GJ 205, GJ 887, GJ their immediate circumstellar environ-
191 (also known as ment are on the order of a few milliarc-
“Kapteyn's star”) and Proxi-
ma Centauri (red filled cir- seconds (mas), inaccessible at the diffrac-
cles; with error bars). For tion limit of current single telescopes but
comparison, the masses and easily resolvable with long-baseline inter-
radii of the Solar System
ferometry. Stellar angular diameters, com-
planets (blue triangles) and
of HD 209458 B (black trian- bined with measurements of bolometric
gle) are also shown. The two fluxes, give model-independent effective
curves represent theoretical temperatures. An accuracy of 50 K or less
models for stars of two dif-
ferent ages (400 million in Teff is required, for example, to put sig-
years - red dashed curve; 5 nificant constraints on the current theo-
billion years - black fully retical models of late-type stars. This cor-
drawn curve). The VLTI responds to an accuracy of 2% or better
results are from Sègransan
et al. (2003), the models in the angular diameter.
from Baraffe et al. (1998). The VLTI routinely achieves a preci-
sion well below the 1% level, and is ideal-
ly suited to expand our knowledge in this
area, especially for main sequence stars
the VLTI, with several students choosing FUNDAMENTAL STELLAR PROPERTIES where the data are very scarce. When
it as the basis of their scientific career in If astronomy has a peculiarity among combined with accurate parallaxes such
their thesis and dissertations. modern scientific disciplines, it must be as those from HIPPARCOS, angular
Amidst all this enthusiastic flowering, that it allows us to obtain insight and diameters can be converted into linear
more than a few remained skeptical. Was understanding of objects that we cannot diameters. Finally, when applied to binary
interferometry really worth the effort? ever subject to our direct scrutiny. From stars, interferometry can measure the
Where was the long-awaited scientific stars to nebulae, from black holes to orbital motions in close pairs: together
production? When would we see the first active galaxies, we predict birth, evolution with spectroscopic radial velocities, this
papers? Solid questions, which have now and death of a universe where distances information leads to the determination of
received a solid answer. This year, the first are so immense that they defeat imagina- the masses and distances.
publications have appeared, and not just tion. Come to think of it, the wonderful Angular diameters have traditionally
one or two as one might expect from a monument of our scientific knowledge constituted the primary targets of all
facility which after all is just emerging rests on a system of pillars which is, after interferometers, and at least in this
from commissioning of its basic systems. all, relatively simple: we have to build on respect the VLTI is no exception. Already
In 2003, more than a dozen papers based fundamental blocks such as stellar mass- from the first observations on both
on VLTI data have been submitted or es, sizes, luminosities and chemical com- siderostats and UTs, several tens of stars
accepted by refereed journals with a sim- position in order to derive almost every- have been observed with diameters sus-
ilar volume of contributions to workshops thing else, from the properties of the ceptible to being resolved on the avail-
and conferences of a scientific nature. newly found extrasolar planets to the able baselines. Faithful to the commis-
The key to this early successful energy liberated at the heart of the most sioning mission, these stars encompassed
exploitation has been the public release distant quasars. a wide variety of types: from calibrator
of all on-sky commissioning data of scien- Interferometry is the key technique to objects with a well-estimated diameter, to
tific interest. Every few months, a new measure some of these fundamental late-type giant stars, to solar analogues
release appears on the VLTI web page, blocks. A recent review of its technical like α Centauri, to pulsating variables
listed by object observed as well as by and scientific aspects provided by Mon- such as Miras and AGB stars. For many of
night of observation. The data can be nier (2003) could be a helpful introduc- these objects, the VLTI measurements
requested from the ESO archive and are
usually delivered very quickly. With much
of the commissioning tasks completed,
the VLTI is beginning to be accessible to
the community for open time observa-
tions. Already for period 73, 30 proposals
for the MIDI instrument have been sub-
mitted on the basis of the standard
ESOFORM package, and they will be
reviewed by the OPC as any other pro-
posal for ESO telescopes. As for all ESO
instruments, a complex and extensive sys-
Figure 2: The correlation meas-
tem of user support, data quality, and
urements obtained on Fomalhaut
archiving has been developed also for the on the night of October 20, 2002,
VLTI. Interferometry for all astronomers as a function of hour angle, i.e. of
might have been a vision until now, but it baseline projected length. The
solid line is the best fit for a uni-
has definitely become a reality on
form-disc angular diameter of
Paranal. 2.086 mas (Davis et al. 2003).

© ESO - December 2003 27

 we imagine the star as a disc of uniform
brightness, but when the accuracy
increases, then we must take into account
phenomena such as limb-darkening.
On both counts, significant help
comes from the VLTI commissioning
efforts. An intense program of observa-
tions of calibrator stars coordinated by A.
Richichi, I. Percheron and M. Wittkowski
has been given high priority. An initial list
of bright candidate calibrators for the
VINCI instrument has been established
Figure 3: Measured squared visibility amplitudes of Psi Phe together with predictions by a (solid based on both estimated diameters and
black line) spherical PHOENIX model atmosphere, (dashed-dotted line) plane-parallel PHOENIX existing previous measurements. A simi-
model atmosphere, (dotted line) plane-parallel ATLAS 12 model atmosphere, (dashed line) lar effort has been undertaken in parallel
plane-parallel ATLAS 9 model atmosphere. The models were constructed by comparison to
spectrophotometry. The gray lines denote uniform disc and fully darkened disc models. The left by the MIDI consortium for the specific
panel shows the full range of the visibility function while the right panel is an enlargement of the needs of their instrument. A dedicated
low squared visibility amplitudes in the second lobe. From Wittkowski et al. (2003). mini-workshop was held in Garching in
January 2003 with about 30 attendees
from ESO, the instrument consortia, and
the community in general.
were the first ever obtained, given that different ages – calculated before the As a result, a list of potential VLTI
the southern sky has remained so far interferometric data became available calibrators has been made public, includ-
largely untapped by interferometers as (Baraffe et al. 2003). ing almost 600 objects for both the
sensitive as the VLTI. These VLTI results show that the VINCI and the MIDI instruments. Up to
A notable achievement of the VLTI in models are at present satisfactory but also August 2003, 133 sources from this list
this field has been the measurement of that new challenges lie ahead. One might have been observed successfully with
the angular diameter of several main notice in Fig. 1, for example, the apparent VINCI. Fifty-two of these objects have
sequence late-type stars, with spectral discrepancy of HD 209458 B from the more than 100 observations each on more
types between M0.5 and M5.5 model. Furthermore, it can be appreciated than 3 baselines. The data are being ana-
(Sègransan et al. 2003). Other main that, for masses around 0.1 MA, the rela- lyzed by the pipeline and by specially
sequence stars of earlier spectral types tion between mass and luminosity is pre- developed global fitting programs aimed
have also been measured by the VLTI: dicted to be quite flat and improved accu- at computing the best solution to all cali-
these include α Centauri A and B, Sirius, racies will be required. If this challenge brator observations on a nightly basis as
α Eridani, Procyon and τ Ceti. As can be met, the reward will consist in well as over several nights (Percheron et
opposed to giant stars, for which hun- strong observational constraints on both al. 2003).
dreds of direct angular diameter determi- atmosphere and interior physics. The While this approach has, by necessity,
nations exist by several techniques mass range between 0.001 and 0.1 MA is the style of an automated large-scale pro-
(Richichi & Percheron 2002), only a particularly interesting since this is the cessing, individual calibrator stars are also
handful of angular diameters are avail- realm of brown dwarfs and eventually of being measured accurately on a case by
able for main sequence stars, which are large planets. Objects in this range are too case basis. An example of this is given by
two orders of magnitude smaller. In par- faint at the moment for the VLTI but will the observations of Fomalhaut (α PsA), a
ticular for the coolest types, the VLTI become accessible with off-axis phase- bright star which, with its small angular
result by Sègransan and his collaborators referencing provided by PRIMA. diameter, lends itself well to being a cali-
has already almost doubled the available The VLTI equipped with the VINCI brator for the hectometric VLTI base-
statistics and the prospects look very test instrument has demonstrated an lines. During continued observations of
promising with the forthcoming introduc- intrinsic accuracy well below the 1% this star over one night, J. Davis et al. fol-
tion of the AMBER instrument. level, and even higher standards are lowed the relative change of the fringe
The conversion from angular diame- expected with the AMBER instrument. contrast as a function of time, i.e. of the
ters to linear sizes is straightforward However, the accuracy in the fringe con- change in the projected baseline due to
thanks to the HIPPARCOS parallaxes for trast is not the only factor in the quest for Earth's rotation (see Fig. 2).
these nearby stars. However, the conver- improved stellar diameters. On one hand, This particular kind of measurement
sion to masses is more problematic, since it is necessary to compare the fringe con- does not require an external calibration,
these stars are single and no direct deter- trast of the science target with that of a provided that the baseline is accurately
mination is possible. Some empirical cali- calibrator star. In this case, the aim is to known. An uncertainty term is introduced
bration of the mass-radius relation exists make sure that the calibrator diameter is by the changing characteristics of the
for the lower main sequence (Delfosse et so precisely known that no additional atmosphere which affects the interfero-
al. 2000), and can be used to convert the error is introduced but, unfortunately, this metric contrast (particularly since a vari-
VLTI measurements into masses. A plot is often not the case at the level of preci- ety of zenith angles are sampled). Davis
of this result as a function of the luminos- sion attained by the VLTI. On the other and collaborators, even after allowing a
ity is shown in Fig. 1, together with the hand, the angular diameter, be it of a ten- large increase in the error bars to take
Solar System planets and the eclipsing uous giant or of a compact main sequence into account such adverse factors,
planet HD 209458 B. The two curves rep- star, can only be as precise as our under- obtained a precise limb-darkened angular
resent theoretical models for stars of two standing of the stellar surface. Generally, diameter for Fomalhaut of 2.109 w

28 The Messenger 114

 data. At the end of their analysis, the
authors were able to determine a thick-
ness of 6% of the atmospheric layer for
this star, and derive a brand new set of its
fundamental properties: Teff = 3472 w
125 K, radius = 86 w 3 RA, mass = 1.3 w
0.2 MA, log g = 0.68. These data constitute
a valuable observational check for theo-
retical model predictions.
Clearly new similar measurements are
needed to constrain the models but this
should be no difficult task for the VLTI,
with a large number of potential candi-
dates available. These measurements con-
stitute also an important precursor for
studying stellar surface structure in more
detail, including features such as surface
spots. Studies of stellar surface structure
require high-precision measurements of
low-contrast fringes in the second and fol-
lowing lobes. Wittkowski et al. (2002)
showed that a realistically modeled spot
on a magnetically active star can likely be
directly detected and its parameters con-
strained using squared visibility ampli-
tudes and triple products obtained with
the VLTI and the AMBER instrument.
Figure 4: Overview of the α Cen A (red) and B (blue) squared visibilities and the corresponding An important new trend in the
best fit uniform disc (UD) models. Details of the lower visibility points are shown in the upper
right (α Cen B) and lower left panels (α Cen A). The dashed lines represent the limits of the W 1σ exploitation of accurate stellar diameters
error. From Kervella et al. (2003). is represented by asteroseismology, and
also in this area the VLTI has started off
with important contributions. The outer
0.013 mas, which confirms and improves strength of the limb-darkening effect for layers of most stars, including our Sun, are
the original measurement of this star by this star has been obtained by Wittkowski constituted by ionised gas which is held in
the intensity interferometer (2.10 w 0.14 et al. (2003). The models were independ- balance between the pull of gravity and
mas, Hanbury-Brown et al. 1974). Consid- ently constructed by comparison to avail- the pressure of the enormous radiation
ering the recognized accuracy of that spe- able spectrophotometry. They used a grid emitted from the interior. These layers
cial interferometer, this is no small prize of stellar atmospheric models, and com- have their own characteristic frequency of
for the VLTI. In parallel, also an puted the corresponding visibilities, oscillation under the influence of these
improved value of the effective tempera- selecting the ones which best fit the VLTI forces and this frequency can be excited
ture Teff=8819w67 K was obtained. data (see Fig. 3). by mechanisms such as convection. As in
Regarding limb-darkening, this This is a delicate process, which must a giant musical bell agitated by an invisi-
important aspect of the stellar surface is take into account issues of calibration and ble hand, the outer layers of these stars
quite difficult to measure directly. The particularly the effect of the broad-band resonate in harmony, resulting in slow but
main reason for this is that the change in filter employed in VINCI, since visibility regular shifts of their spectra.
the visibility curve introduced by limb- functions are strongly chromatic and the These effects have been very well
darkening, as opposed to a uniform or effect of limb-darkening varies across the identified in our Sun, and are now begin-
fully darkened disc, is noticeable mostly band. Even several definitions of the ning to be measured in some of the stars
at the baseline frequencies after the first angular diameter exist, and one must take closest to us. In fact, theory predicts a
minimum, and then only marginally (see care to be consistent between models and whole system of characteristic frequen-
Fig. 3). As a consequence, very few meas-
urements are available, but the VLTI is
moving quickly to fill this gap. Already
only one month after the first UT fringes,
Figure 5: Location of τ Cet in
commissioning observations of the M4 the Hertzsprung-Russell dia-
giant star Psi Phe were obtained. Using gram, according to the deter-
the large sensitivity of these telescopes, as mination inferred from inter-
ferometry (left) and spec-
well as their long baseline, the visibility troscopy (right), with the
was accurately measured in the critical associated uncertainty boxes.
range beyond the first minimum. By com- The solid and dashed lines
bining these measurements with others at correspond to two models of
the star, with significantly dif-
lower frequencies (for which the ferent values of the scale of
siderostats were sufficient), an accurate the mixing-length in the star.
confirmation of the model-predicted From Pijpers et al. (2003).

Richichi A. & Paresce F., Harvesting Scientific Results with the VLTI © ESO - December 2003 29
 cies for each star, depending on mass,
radius, temperature of the gas. Most
importantly, they depend on details of
their interior composition and energy
production mechanisms. Like a finger-
print, the oscillation frequencies of each
star are different from any other. Unfor-
tunately, the number of parameters
involved is so large that observations are
often well consistent with several, signifi-
cantly different models.
Now, long-baseline interferometry is
coming to the rescue, and once again with
its most basic and simplest type of meas-
urement: the accurate determination of a
stellar angular diameter. At least for near-
by stars where accurate parallaxes are
available, we can, thus, obtain the linear
size and this is a precious piece of infor-
mation for asteroseismologic modeling. Figure 6: VLTI ground baselines for Achernar observations and their corresponding projections

In its debut year, the VLTI has already onto the sky at different observing times. Left, scheme of VLTI baselines for the two pairs of
siderostats used for Achernar observations. Colour magenta represents the 66 m (E0-G1;
produced four papers with this important azimuth 147-, counted from North to East) and green the 140 m (B3-M0; 58-). Right, Corre-
keyword. In some cases, the stars meas- sponding baseline projections onto the sky (Bproj ) as seen from the star. Note the very efficient
ured are very well known like α Cen, Sir- Earth-rotation synthesis resulting in a nearly complete coverage in azimuth angles.
ius, and Procyon, and the corresponding
results are certainly exciting for experts
and casual readers alike. G8, temperature approximately 5300 K). cantly different assumptions on the initial
We take here as illustrations of the More interestingly, while the direct meas- hydrogen content and the scale of the
VLTI contribution to this field, the well urement by the VLTI resulted in a diame- mixing-length. Further observations by
known stars α Cen A and B (Kervella et ter largely in agreement with theory in the VLTI, along with a more refined data
al., 2003) our closest stellar neighbour the case of α Cen, in the case of τ Cet analysis, are certainly desirable.
and the lesser known star τ Cet (Pijpers et there appears to be a significant differ- Asteroseismology predicts oscillation
al. 2003). The angular diameters of the ence (see Fig. 5). The radius and effective frequencies quite different for the two sit-
two main components of the system, α temperature of this star estimated on the uations above: a peak at 3570 µHz and a
Cen A and B, were measured using basis of its spectral and photometric char- frequency spacing of 173 µHz in the case
VINCI with a relative precision of 0.2% acteristics are 0.87 W 0.04 RA and 5264 W favoured by the interferometric measure-
and 0.6% respectively. The measured uni- 100 K respectively, while using VLTI data, ment, and 2950 µHz and 1148 µHz,
form disc angular diameters for α CenA Pijpers et al. (2003) obtained 0.773W respectively for the other case. High-
and B were 8.314 W 0.016 and 5.856 W 0.004 RA and 5525W12 K. accuracy spectroscopic measurements
0.027 mas, respectively and limb dark- However, they also recognize that with an instrument like HARPS will be
ened angular diameters of 8.511 W 0.020 their estimate suffers from having only able soon to shed light on this discrepan-
and 6.001 W 0.034 mas, respectively (see one single calibrator available, and that cy. In the future, high-accuracy interfero-
Figure 4). the actual uncertainty could be five times metric and spectroscopic measurements
Particular care was taken in the cali- larger than the formal one. The difference will go hand in hand, and allow us to
bration of these measurements consider- between these results is shown in Fig. 5, understand better the internal composi-
ing that VINCI estimates the fringe visi- which makes clear that the two sets of val- tion and energy transfer mechanisms of
bility using a broadband K filter. Combin- ues correspond to models with signifi- many more stars.
ing these values with the known parallax,
the linear diameters of 1.224W 0.003 DA
and 0.863 W 0.005 DA were derived for
the two components A and B, respective-
ly. The measurement of α Cen A is the
most precise photospheric angular size
ever obtained by interferometry.
The measurements were compared to
Figure 7: Fit of an ellipse over the observed
recent model diameters constrained by V 2 points of Achernar, translated to equiva-
asteroseismic observations. The reported lent uniform disc angular diameters. Magen-
values are compatible with the most ta points are for the 66 m baseline and yel-
low points are for the 140 m baseline. The
recently published masses for both stars. fitted ellipse results in major axis 2a= 2.53 W
If α Cen, the closest solar analogue, is 0.06 milliarcsec, minor axis 2b= 1.62 W 0.01
slightly hotter and larger than our Sun, τ milliarcsec, and minor-axis orientation
Cet provides an interesting bracketing α0=39-W1- (from North to East). The points
distribution reveals an extremely oblate
comparison because it is slightly smaller shape with a ratio 2a/2b = 1.56 W 0.05.
and cooler than our Sun (spectral type From Domiciano de Souza et al. (2003).

30 The Messenger 114

 NEW ASPECTS OF STELLAR
RESEARCH: ACHERNAR
AND ETA CARINAE
Stellar angular diameters may be consid-
ered the staple food of interferometry, but
as sensitivity and accuracy increase new
exciting results are produced even with
this basic kind of measurement. We have
seen that already in the case of more or
less regular stars such as late-type giants
and solar analogues, we are beginning to
tackle interesting subjects such as limb-
darkening and asteroseismology. Howev-
er, when we move to less ordinary objects,
interferometry is the key to open com-
pletely new doors. The VLTI is bringing
facts, after many promises, also in this
area, and as an example we take the
recent results obtained on the fast rotat-
ing star Achernar, and the ultra-luminous
star η Carinae.
Rotation is an intrinsic property of all
stars, but for some classes of stars it can
be a rather extreme phenomenon, with
important consequences on the stellar
structure. The most obvious is the geo-
metrical deformation that results in a
radius larger at the equator than at the
poles. Another well established effect,
known as gravity darkening or the von
Zeipel effect, is that both the surface
gravity and emitted flux decrease from
the poles to the equator. Although well
studied in the literature, such effects of
rotation have rarely been directly tested
against observations. The best candidates
for such observational tests are represent-
ed by Be stars. A Be star is defined as a B Figure 8 : Zoom into the η Carinae nebula. Top left: WFPC2 image (Morse et al. 1998). Top right:
type star that has presented episodic NACO observations at 2 µm. Centre left : VINCI data reveal an object with size 5 mas. This is
Balmer lines in emission, whose origin is not the photosphere of the star, but the radius at which the stellar wind becomes opaque. Bot-
tom left: VINCI data, converted to an effective diameter, plotted against the position angle of
attributed to a circumstellar envelope the baseline. Bottom right: the diameter change with position angle implies that the object is
(CSE) ejected by the star itself. Physical elongated; the orientation is the same as that of the large-scale nebula shown in the top left
mechanisms like non-radial pulsations, panel. Courtesy R. Van Boekel.
magnetic activity, or binarity have been
invoked to explain CSE formation in Be
stars in combination with their funda-
mental property of rapid rotation. ical velocity or not. This will have a pro- out from 11 September to 21 December
Struve’s original vision of a critically found impact on the dynamical models 2002, with quasi-uniform time coverage
rotating Roche star ejecting material for Be star disc formation due to rapid using the siderostats. Two interferometric
from its equator has been discarded in the rotation combined with mechanisms like baselines (66 m and 140 m) were used
past by observing that Be stars rotate at pulsation, radiation pressure of photos- (Fig. 6 left). Their orientations are almost
most at 70% to 80% of their critical pheric hot spots or expelled plasma by perpendicular to each other giving an
velocity (typically 500 km/s for a B0V magnetic flares. excellent configuration for the detection
star) a value not sufficient to explain the The southern star Achernar (α Eri- of stellar asymmetries. Moreover, Earth
presence of discs. However, this statisti- dani, spectral type B3Vpe) is the brightest rotation produced an efficient synthesis
cally observed limit may be biased by the Be star in the sky and, therefore, a perfect effect (Fig. 6 right).
fact that close to or beyond such veloci- target for the VLTI and the siderostats. It Analysis of the processed data gives
ties the diagnosis of Doppler-broadened also represents a convenient object to test the results summarized in Fig.7 which
spectral lines fails to determine the rota- the validity of the concepts briefly reveal an extremely oblate shape from
tion value due to gravity darkening. Only described above. A recent paper by the distribution of equivalent UD diame-
direct measures of Be star photospheres Domiciano de Souza et al. (2003) ter values on an ellipse. The results of this
by interferometry can overcome the chal- describes VLTI observations of this star fit are: major axis 2a=2.53W0.06 mas,
lenge of proving whether these objects and exciting new results. Dedicated minor axis 2b=1.62W0.01 mas and minor-
rotate close to a few percent of their crit- observations of Achernar were carried axis orientation α0=39-W1-. Note that the

Richichi A. & Paresce F., Harvesting Scientific Results with the VLTI © ESO - December 2003 31
 Figure 9: Left: 3.4 u 3.4 arcmin optical image of NGC 1068, (NOAO/AURA/NSF). Centre: non-interferometric acquisition image of NGC 1068
taken by MIDI with a 8.7 micron filter, showing the structures on arcsec scales. Also shown are the position of the spectroscopic slit used in the
interferometric observations and the directions of North (toward top left) and East (toward bottom left) on the sky. The projected baseline was
essentially North/South and the fringe spacing in this direction was 26.3 mas at 10 micron wavelength. Right: sketch of the dust structure in the
nucleus of NGC 1068, as derived from modeling the MIDI observations. It contains a central hot component (T > 800 K, yellow) which is signif-
icantly smaller than the interferometric beam, and a much-larger well-resolved warm component (T=330 K, red) of diameter 33W5 mas, corre-
sponding to 2.8 pc at the distance of NGC 1068. From Jaffe et al (2003).

corresponding ratio 2a/2b=1.56W0.05 been between 150 and 200 MA. It loses A second important conclusion from
determines the equivalent star oblateness mass at a prodigious rate (of order 10−3 the VLTI data is that the central object is
only in a first-order UD approximation. It MA/year) in a 500 km/s wind. In an enor- not spherically symmetric (Figure 8, cen-
can be shown that, in the particular case mous eruption in the middle of the 19th tre left panel). In fact, its major axis is
of Achernar, the observed asymmetry of century, several solar masses were eject- aligned with that of the large-scale struc-
Achernar reflects its true photospheric ed; the nature of this eruption is still not ture (Figure 8, bottom right panel). This
distortion with a negligible CSE contribu- understood. The resultant debris now alignment on all scales means that the
tion. Under this assumption, and using forms a large prolate nebulosity sur- 1840 outburst looks like a scaled-up ver-
the Hipparcos distance (d=143.8W3.6 rounding the star, with an elongation sion of the present-day wind, and that this
light-years), an equatorial radius Req = along a position angle of 135-. Clumps are wind is stronger along the poles than in
12.0 W 0.4 RA and a maximum polar found at all spatial scales; the strong inho- the equatorial plane. This can be under-
radius equal to 7.7 W 0.2 RA can be mogeneities make it impossible to deter- stood in the framework of radiation-driv-
derived from the equivalent UD meas- mine the mass loss rate from spectroscopy en winds from rapidly rotating stars: cen-
urements. From simple geometrical con- alone. trifugal forces favour mass-loss in the
siderations, the actual polar radius Rpol The highest-resolution observations equatorial plane, but the radiation pres-
will be smaller than for polar inclinations of η Car from a single telescope are the sure in these massive stars is stronger in
i < 90-, while Req is independent of i. VLT / NACO data, shown in the top right the polar regions because of the von
Using an interferometry oriented panel of Figure 8. They resolve much of Zeipel effect (the stronger gravity near
code which includes radiation transfer the sub-arc second structure, but about the poles leads to a higher temperature).
and the von Zeipel law applied to Ach- 60% of the flux within the inner 1.5) For η Car, the von Zeipel effect is
ernar in the Roche approximation, it was remains unresolved in a central object more important than the centrifugal levi-
found that the commonly adopted Roche whose size must be smaller than 70 mas. tation leading to a polar wind. The VLTI
approximation (uniform rotation and VLTI / VINCI observations clearly observations, thus, favour a model that
centrally condensed mass) fails to explain resolve this central object; its size can now interprets the morphology of η Car on all
Achernar's extreme oblateness. This be measured to be 5 mas at 2 µm corre- scales with a radiation-driven wind from a
result opens new perspectives in basic sponding to 10 AU at η Car’s distance. rapidly rotating star, and they have
problems in stellar physics such as rota- This is clearly much larger than the stellar allowed us to more precisely determine
tionally enhanced mass loss of Early-type photosphere so that we must be observ- the very high mass-loss rate from this
stars. In addition to its intimate relation ing the radius at which the stellar wind object.
with magnetism and pulsation, rapid rota- becomes opaque. The radiation is domi-
tion thus provides a key to understanding nated by free-free emission and electron A NEW WINDOW IS OPEN: INTERFER-
the Be phenomenon, which is one of the scattering; the radius of the surface is OMETRY OF EXTRAGALACTIC OBJECTS
outstanding non-resolved problems in determined by the mass-loss rate and the The few selected results presented so far
theoretical astrophysics. wind clumping factor. The diameter meas- are just highlights of the massive amount
If energetic rotation is the key for a urement with the VLTI breaks the degen- of observations accumulated by VINCI,
star such as Achernar, pure energy is the eracy between these two parameters in the test instrument which was originally
main keyword for the next object in the previous modeling efforts; mass loss rate designed to test the VLTI at system level,
new flavour of angular diameters that the and clumping factor can be derived sepa- and has instead operated almost continu-
VLTI is biting into. With 5 N 106 LA, η rately from the combination of HST / ously for over two years. Soon, the two
Carinae is the most luminous star known STIS spectroscopy with the interferomet- facility instruments MIDI and AMBER
in the Galaxy; its initial mass must have ric data. will enable observations with an

32 The Messenger 114

increased range of wavelengths, spectral ed an historical landmark for the whole component is located in front of the hot
dispersions and number of beam combi- field of interferometry, often regarded as component. The narrow versus wide
nations. In fact, MIDI arrived on Paranal a tool useful only for stellar research. hatching in Fig. 9 indicates that the sili-
almost exactly one year ago, and having Even more so, when one considers that cate absorption towards the hot compo-
successfully achieved first fringes as well almost simultaneously the Keck Interfer- nent is significantly larger than the (aver-
as having undergone a few commission- ometer, making use of their two 10 m tel- aged) silicate absorption towards the
ing runs, it is now offered for open obser- escopes equipped with adaptive optics, warm dust. Although the observations
vations in period 73 starting from April also obtained for the first time an inter- clearly resolve the warm component
2004. In the meantime, under a shared- ferometric observation of NGC 4151, an along PA 58-, there is no spatial informa-
risk basis, already some GTO and science extragalactic source, in the near-infrared. tion available in the orthogonal direction:
demonstration observations have taken Both galaxies belong to the well- this uncertainty is indicated by the dashed
place. known category of active galactic nuclei ellipse. The displacement of the warm
The results are certainly impressive: (AGN), some of the most spectacular component by several mas relative to the
from young stars, to evolved late-type objects in the sky. AGNs display a pletho- hot core indicates that the current inter-
objects, from the luminous star η Car pre- ra of very energetic phenomena: relativis- ferometric data allow for both compo-
viously described, to early-type emission tic jets, broad and narrow emission lines, nents being not concentric.
line stars, dozens of targets have been X-ray continuum and line emission, radio These MIDI observations represent
observed. One should note that MIDI lobes all of which probably ultimately the first interferometric spectrum of an
represents the first Michelson-type beam originate from the accretion of matter extragalactic source at IR wavelengths,
combiner operating at 10 microns ever to onto a central supermassive black hole. and they reveal a deep absorption feature
scan the skies in a routine fashion (ISI, The varying relative importance of these which appears significantly different from
another innovative and successful inter- phenomena results in a complex classifi- that seen in our galaxy. The 10 micron fea-
ferometer operating at these wavelengths cation scheme which includes quasars, ture revealed in the interferometric spec-
is based on a different principle of beam radio galaxies, BL Lac objects, Seyfert 1 trum provides, at a first interpretation,
combination using heterodyne technolo- and 2 galaxies etc. It is generally agreed evidence for an alumino-silicate composi-
gy). Add to that the fact that it is fed by that at least part of the observed diversity tion of the central dust (see Fig. 10). This
the giant 8.2 m mirrors of the VLT, and it is caused by orientation effects: from cer- is different from the olivine-type dust
comes as little surprise that even these tain viewpoints, circum-nuclear dust (commonly seen around stars in our
initial observations are rich in new results. blocks the direct view of the central Galaxy) feature that is revealed in non-
Indeed, every object that MIDI is pointed accretion disc and central jets. However, interferometric, single telescope spectra.
towards represents a first timer for this this model still remains to be demonstrat- However, other more exotic interpreta-
kind of observation. It is certainly not too ed by direct observations, since the angu- tions such as the existence of PAHs com-
optimistic to expect that several new lar resolution provided by even the pounds, cannot be yet completely exclud-
exciting discoveries will be made possible largest telescopes fails to resolve the dust ed. Besides, the observations need to be
soon by MIDI. geometry in even the nearest AGNs. complemented by more baselines to
Already now, new ground has been The first interferometric observation extend the coverage of position angles
broken: the first 10 micron interferomet- of NGC 1068, a Seyfert 2 AGN, by MIDI and resolutions. Such observations are
ric observation of an extragalactic source, at the VLTI has probed the inner regions being carried out as this article is being
NGC 1068. This result was one, but not of this object with the unprecedented res- written, and hopefully we will soon hear
the only, outcome of the first observations olution of 30 mas. The combined spatial more exciting news on this.
in the framework of the science demon- and spectral information reveals that the
stration program. This is a form of guar- central dust distribution has a size of a THE FUTURE, NOT SO FAR
anteed time awarded to those partners few parsecs and contains an unresolved The scientific results described above are
who have made practical contributions to hot core, which might be the outer part of only a selection of what has been
the VLTI, an open club where new mem- the accretion disc (see Fig. 9). achieved so far. Other data already avail-
bers are always welcome. This represent- A significant amount of the warm able are currently being analyzed or wait-
ing for additional complementary obser-
vations by the VLTI as well as by other
instruments. Preliminary results have
Figure 10: The interferometric
already been presented at a number of
spectrum of NGC 1068 showing
the flux on scales of 30 milli-arc- meetings and conferences, from national
sec and smaller. The jagged line and international meetings to the XXV
shows the data and the smooth IAU General Assembly, from topical
line the best fitting two-compo- symposia to the recent MIDI workshop
nent gaussian model described in
Fig. 9. The red and green lines held last September. Astronomers will
indicate the individual contribu- soon be able to make additional impor-
tions of the "hot" and "warm" tant steps in several areas, including stel-
components, respectively. The dip
near 10 micron is probably
lar diameters, pulsation, circumstellar
caused by alumino-silicate dust, environments and binary stars.
as opposed to the olivine-type, Noteworthy are the results expected
silicate dust absorption noted in in the fields of pulsation, both among
the same spectrum obtained
without the interferometric com- Cepheid stars and very evolved objects.
bination. From Jaffe et al. (2003). The former are expected to permit a deci-
sive improvement in the accuracy of the

Richichi A. & Paresce F., Harvesting Scientific Results with the VLTI © ESO - December 2003 33
 empirical period-luminosity relation for added in the underground tunnel: the for- FIRST PUBLICATIONS RELATED
these important distance indicators, with mer will extend the sensitivity by permit- TO THE VLTI:

mass stars with the VLTI, Sègransan

important implications also at an extra- ting long integrations on faint objects, the First radius measurements of very low
galactic level. Regarding the latter, pulsa- latter will enable a much larger set of
tion and mass loss of evolved stars are a baseline combinations than has been pos- et al., 2003, A&A, 397, L5
key to understanding important phenom- sible until now. Finally, the two facility The diameters of α Centauri A and B: a
comparison of the asteroseismic and
ena such as the late-stages of the evolu- instruments AMBER and MIDI will offer
VLTI views, Kervella et al., 2003,
tion of stars like our Sun, and are also a wavelength coverage which spans the J, A&A, 404, 1087
responsible for the chemical enrichment H, K and N bands respectively, with vari- Interferometry and asteroseismology:
of the interstellar medium. Already a ous sets of spectral resolution up to The radius of τ Cet, Pijpers et al.,
large number of observations have been 10,000. 2003, A&A, 406, L15
collected by the VLTI for these kinds of With these additions, the VLTI will The spinning-top Be star Achernar from
objects. soon be a more complete facility, but its VLTI-VINCI, Domiciano de Souza et
This wealth of initial results attests to development will be by no means termi- al., 2003, A&A, 407, L47
The interferometric diameter and inter-
the fact that the VLTI is working well. nated. Second-generation instruments are
nal structure of Sirius A, Kervella et
This might come as no surprise for the being proposed, including the ESA-fund-
al., 2003, A&A, 408, 681
many engineers and astronomers that ed GENIE nulling interferometry First VLTI observations of Mira stars,
dedicated so much effort through the demonstrator for DARWIN. The PRIMA Richichi and Wittkowski, 2003,
years to its design and construction, but it facility, crucial for the observation of faint Ap&SS, 286, 219
certainly is comforting news to the large sources and for accurate narrow band The VLTI Calibrators Program: Getting
community that is longing for long base- astrometry through the use of off-axis ref- ready for high accuracy measure-
line interferometry as a wide-ranging, erence sources, is planned to be integrat- ments, Percheron, Richichi, and
user-friendly and reliable technique to be ed in about two years. It will extend the Wittkowski, 2003, Ap&SS, 286, 105
Direct measurement of the size of the star
used by many new enthusiasts and not sensitivity limit of the VLTI well into the
η Carinae, Van Boekel et al., 2003,
just a few black belts of applied opto- realm of extragalactic sources and permit A&A, in press (astro-ph/0310399)
mechanics. Still, one should not forget the detection of the tiny gravitational Tests of stellar model atmospheres by
that all this wealth comes from what has pulls induced on stars by their orbiting optical interferometry: VLTI/VINCI
been until now a rather limited configura- planets. In the course of 2003, what was limb-darkening measurements of the
tion of the VLTI. The majority of the first once a tiny seed timidly buried in hard M4 giant ψ Phe, Wittkowski et al.,
observations have been limited in sensi- ground, has finally grown and started pro- 2003, A&A, in press
tivity by the size of the small test ducing its fruits. The harvesting has The diameter and evolutionary state of
siderostats, the lack of adaptive optics begun, and the prediction is that it will Procyon A, Kervella et al., 2003,
A&A, in press (astro-ph/0309148)
and of a fringe tracker. Flexibility has last many seasons.
J-K DENIS photometry of a VLTI-select-
been severely limited by the lengthy relo- ed sample of bright southern stars,
cation of the siderostats and by the num- ACKNOWLEDGEMENTS Kimeswenger et al., 2003, A&A, in
ber of delay lines (three) currently If we can speak about VLTI results today, press
installed. Last but not least, the test it is only thanks to the efforts of a large Introduction to VINCI/VLTI interfero-
instrument VINCI is equipped with a group of people in Europe and Chile. It is metric data analysis, Kervella et al.,
impossible to mention here all their names, 2003, A&A, submitted
wide band K filter only.
and we hope we can be forgiven for this Calibration observations of Fomalhaut
All these limitations are being lifted
negligence. However, we take special pleas- with the VLTI, Davis et al., 2003,
as we write. The first two of four MACAO ure in mentioning the former head of the A&A, submitted
adaptive optics systems specific for the Telescope Division, Massimo Tarenghi, Dust in the nucleus of the active galaxy
VLTI have been put into operation on the who has defended the VLTI seed from the NGC1068: structure and composition
UTs, and the rest will follow in the course very first moment. Although he is now on parsec scales, Jaffe et al., 2003,
of next year. In parallel, the first 1.8 m devoting himself to other projects, he still Nature, submitted
Auxiliary Telescope (AT) has been pays the occasional, unannounced sur-
accepted in Europe and is about to start prise visit at the VLTI and still shows the
its test period in Paranal. It will be soon same enthusiasm as ever. We would also OTHER REFERENCES
be joined by the second AT in the course like to mention by name the current mem- Baraffe, Chabrier, Allard, Hauschildt, 1998,
of 2004, while two more will follow later: bers of the VLTI Science Group, on both A&A 337, 403
sides of the Atlantic: Pascal Ballester, Baraffe, Chabrier, Barman, Allard, Hauschildt,
together, the ATs will constitute an inde-
Emmanuel Di Folco, Emmanuel Galliano, 2003, A&A 402, 701
pendent observatory on Paranal, fully
Andreas Glindemann, Christian Hummel, Dwarkadas, Owocki, 2002, ApJ 581, 1337
dedicated to interferometry and quickly Pierre Kervella, Sebastien Morel, Monnier, J.D. 2003, Reports on Progress in
relocatable on a system of 30 stations. The Francesco Paresce, Isabelle Percheron, Physics, Volume 66, Issue 5, p. 789
FINITO fringe tracker is about to com- Fredrik Rantakyrö, Andrea Richichi, Richichi A., Percheron I. 2002, A&A, 386, 492
plete its commissioning on the mountain, Markus Schöller, Martin Vannier, and Wittkowski, Schoeller, Hubrig, Posselt, von der
and three more delay lines are ready to be Markus Wittkowski. Luehe, 2002, AN 323, 241

34 The Messenger 114

 O UR OWN S TARBURS T
IN THIS ARTICLE WE PRESENT THE FIRST RESULTS FROM A NEAR-INFRARED CAMPAIGN TO CHARACTERIZE OUR
GALAXY’S OWN STARBURST EVENT, W49A, A PRODIGIOUS FACTORY OF MASSIVE STARS AT A DISTANCE OF ABOUT
12 KPC AND CONCEALED FROM OBSERVATIONS AT VISIBLE WAVELENGTHS BY MORE THAN 25 MAGNITUDES OF
INTERVENING DUST EXTINCTION. OUR RESULTS SO FAR REVEAL THE PRESENCE OF PREVIOUSLY UNKNOWN
MASSIVE STELLAR CLUSTERS CONTAINING MORE THAN 100 OB STARS, SOME AS MASSIVE AS 120 MA, MOST
STILL EMBEDDED IN THEIR PARENTAL MOLECULAR CLOUD AND WITH AGES AS YOUNG AS 104−5 YR. WE ARGUE THAT
THIS ONGOING STARBURST APPEARS TO HAVE BEEN MULTI-SEEDED INSTEAD OF RESULTING FROM A COHERENT
TRIGGER.

J O Ã O A LVES 1 AND N ICOLE H OMEIER 1,2

1EUROPEAN SOUTHERN OBSERVATORY, GARCHING, GERMANY
2JOHN HOPKINS UNIVERSITY, BALTIMORE, USA

8 MA Well known to radio astronomers, W49A lutionary state. For this reason W49A is

S
TARS WITH MASS ABOVE
are the main suppliers of heavy (Mezger et al. 1967, Shaver et al. 1970) is unique in our Galaxy.
elements in the Universe, the one of the brightest Galactic giant radio In an attempt to uncover and characterize
same elements that make up H II regions (~10 7 LA), powered by the the embedded stellar population in
your body and the Messenger equivalent of about 100 O7 V stars. It is W49A we performed an unbiased 5(×5(
article you are reading. Massive stars also embedded in the densest region of a ~10 6 (16 pc × 16 pc), deep J, H, and Ks-band
inject energy into a galaxy’s interstellar MA Giant Molecular Cloud (GMC) imaging survey with the SOFI camera on
medium (ISM) playing a critical role in extending more than ~100 pc in size ESO’s New Technology Telescope (NTT),
regulating star formation and driving (Simon 2001) and is the best Galactic centred on the densest region of the W49
galaxy evolution. It is not clear today if analogue to the starburst phenomenon Giant Molecular Cloud (Simon 2001)
massive stars are destroyers of embryonic seen in other galaxies. W49A lies essen- (see Figure 1).
planetary systems or if they act, under tially on the Galactic plane (l = 43.17°, b =
some conditions, as triggers to planet for- +0.00°) at a distance of 11.4W1.2 kpc THE FIRST RESULTS
mation. Finally, because they die in spec- (Gwinn 1992), has ~40 well studied ultra- The observations were taken in June
tacularly bright fashion as supernovae/ compact (UC) HII regions (e.g., De Pree 2001, with the SOFI near-infrared camera
gamma-ray bursts, their death can be seen et al. 1997), each associated with at least on the ESO 3.5 m New Technology Tele-
across the Universe providing unique one stars earlier than approximately B3. scope (NTT) on La Silla, Chile, during a
information to observational cosmology. About 12 of these radio sources are spell of good weather and exceptional
Despite their fundamental role in shaping arranged in the well known Welch “ring” seeing (FWHM ~ 0.5)) (see Alves &
our Universe, our knowledge of how (Welch et al. 1987). A few other young Homeier 2003). Follow-up Adaptive
Nature forms massive stars is rather prim- Galactic clusters have a large number of Optics observations of an embedded
itive, in part because objects for study are massive stars, e.g., η Carinae, NGC 3603, compact HII region powered by a new-
rare due to the combination of small the Arches cluster, but no other known born 80 MA candidate star were taken
number statistics and the rapidity with region has a high number of massive stars with NACO on the Very Large Telescope
which they pass through their early in such a highly embedded and early evo- (VLT) during September 2003.
stages. With this in mind, the abundance
of embedded massive stars in the Galactic
star-forming region W49A marks it as a
scientific gem.

Figure 1:
13CO Map of W49 giant molecu-

lar cloud. Our NTT-SOFI survey, marked as

a red box, targeted the densest regions of
this 106 MA giant molecular cloud. The sur-
vey covers an area of 16 u 16 pc at the
distance to the complex (11.4 w 1.2 kpc).
13CO data taken from Simon et al. (2001).

© ESO - December 2003 35

 ing that the Ks extended emission is dom-
inated by hydrogen lines. The most
prominent are identified in Figure 2 fol-
lowing De Pree (1997) nomenclature.
Several point sources lie prominently in
the centre of some of these regions (e.g.,
CC, O3, W49A south), are extincted by
AV > 24 mags of visual extinction (see
below), and are excellent candidates to be
the exciting sources powering these
regions (Homeier & Alves, 2004).
The main feature in Figures 2 and 3 is the
central 6 pc diameter region E of the ring
of radio sources, with a stellar cluster at
its projected centre. From here on we will
refer to this cluster as Cluster 1. Note that
only the North part of this 6 pc region is
visible in the JHKs colour composite, sug-
gesting that there is a larger optical depth
towards the South of Cluster 1, perhaps
due to chance alignment of the embedded
compact regions (e.g., JJ, O3) in front of
it.
We present in Figure 4 the spatial distri-
bution of the detected sources as a func-
tion of (H-Ks) colour. In Figure 5 we pres-
ent the H-Ks vs. Ks Colour–Magnitude
diagram for our survey. The solid line rep-
resents a 1 Myr old population taken
from the Geneva tracks (Lejeune &
Schaerer, 2001) and the slanted dotted
Figure 2: 5(u 5( JHK colour composite of our survey. North is up and East is to the left. The
lines represent a reddening in this dia-
s
red, green, and blue channels are mapped logarithmically to the Ks, H, and J-band respective- gram of AV = 48 mag. The filled circles
ly. The labels identify known radio continuum sources (De Pree et al., 1997). Sources F and J2 represent sources likely associated with
are UC HII regions in the Welch ring. The main cluster (Cluster 1) is seen NE of O3. Several can- the new clusters (see Figure 4).
didate exciting sources of compact HII regions are visible (e.g., the sources at the centre of the
CC, O3, W49A South HII regions). Dark pillars of molecular material are seen associated with
radio sources Q and W49A South. None of W49A sources are optically visible. The coordinates UNCOVERING THE BEAST
of the image centre are: 19:10:16.724; +09:06:11.16 (J2000). Since the W49A star forming region is at
a distance of 11.4 kpc, virtually on the
Galactic plane, one expects a large
In Figure 2 we present the composite taken from De Pree et al. (1997) and has amount of unrelated line-of-sight dust
JHKs colour image for our NTT-SOFI a spatial resolution of 0.8), close to the extinction to W49A, as well as dust asso-
survey. The image covers an area of 5(u5( spatial resolution of the NTT images. The ciated with the star forming region. We
on the sky and the red, green, and blue red-only features in this image represent will take advantage of the large amounts
channels are mapped logarithmically to regions of ionised hydrogen so deeply of dust extinction to isolate a reliable stel-
the Ks, H, and J-band respectively. embedded in the W49A molecular cloud lar population associated with the W49A
Because most field stars are essentially that they cannot be detected in our Ks- giant molecular cloud. We present in Fig-
colourless in the near-infrared one band image, e.g., the Welch ring of UC ure 4 the spatial distribution of the
expects the colour of a star in this image HII regions with the exception of sources detected sources as a function of (H-Ks)
to be, to first order, a qualitative measure F and J2 (that appear
of the amount of extinction towards this in yellow in the
star. For this reason, all blue stars in this image). Several HII
Figure are foreground sources to the star regions and UC HII
forming region. regions detected at
In Figure 3 we present a 3.6 cm radio con- radio continuum
tinuum (red), Ks (green), and J (blue) wavelengths are
colour composite of the central regions of clearly detected on
the survey. The radio continuum data is the Ks-band, suggest-

Figure 3: J (blue), K (green), and 3.6 cm radio continuum (red) colour

s
composite of W49A star forming region. Red only features represent
radio HII regions too embedded to be detected in our deep Ks image
(e.g., most of the Welch ring of UC HII regions). Yellow represents fea-
tures seen in both the radio continuum and Ks-band, green-only
sources are Ks sources too reddened to be detected at the J-band
(essentially all of W49A young stellar population), while blue sources are
foreground stars unrelated to the star forming region.

36 The Messenger 114

Figure 4: Spatial distribution of detected surprisingly well with the luminosity of easily imagine that, chronologically, the
sources as a function of (H-Ks) colour. The the entire region (about 10 51 Lyman con- densest part of the W49A GMC collapsed
clusters (labeled 1, 2, 3, and 4) become tinuum photons emitted per second). We to form Cluster 1, and the combined
apparent in the third panel (where AV ~ 20
mag). The non-uniform distribution of sources can now say that the census of massive action of stellar winds and UV radiation
in panel 1 and 2 could be due to the interven- stars in W49A agrees with the number of compressed the abundant nearby gas to
ing cloud GRSMC 43.30-0.33 located at a Lyman continuum photons derived from the West, triggering its collapse into the
distance of ~3 kpc. The field shown is the radio observations. In fact, photon leak- Welch ring. Recent theoretical calcula-
same as in Figure 1.
age or absorption by dust could be oper- tions suggest timescales for massive star
ating in the W49A region, as our count of formation of the order of ~10 5 yr which
colour. Starting with the bluer sources, candidate O-stars (incomplete due to for a typical sound speed of ~10 km/s
(H-Ks < 1 mag) we find a non-uniform severe inhomoge-
distribution in which about two third of neous extinction), Table 1: W49 Stellar Clusters
the sources are found on the southern added to the num-
ber of known UC Cluster R.A. (J2000) Decl. Associated
half of the field. Sources in this first bin Radio Source
are mainly foreground sources to the HII regions, gives
W49A star forming region, extincted by ~140 stars with
1 19h 10m 17. 5 s +9! 06( 21) Extended
less than about 14 magnitudes of visual masses greater 2 19h 10m 21.9s +9! 05( 04) W49 A south
extinction, and the non-uniform distribu- than 15−20 MA, 3 19h 10m 11.9 s +9! 05( 28) S
tion is likely caused by an intervening suggestive per- 4 19h 10m 10.8 s +9! 05( 14) Q
cloud at a distance of about 3 kpc (cloud haps of a slight
GRSMC 43.30−0.33, Simon et al. 2001). overabundance of
The second panel (1 < H-Ks < 1.5 mag; ionizing stars. An estimate of the Initial agrees well with the crossing time in the
14 < AV < 24 mag) further suggests this Mass Function (IMF) in the main cluster Welch ring. However, we argue that the
interpretation. We see the opposite spa- suggests a Salpeter IMF down to ~ 2− formation of the three smaller clusters to
tial distribution with an increase in 3 MA. the South, undoubtedly associated with
extinction and the region that in the first It is remarkable that the Welch ring of the burst of star formation in W49A, is
bin seemed under-populated is now over- UC HII regions is seen in projection unlikely to have been triggered by Cluster
populated. The majority of these stars are against the edge of the giant region pow- 1. The minimum distance (because of pro-
likely to be highly reddeneded stars in the ered by Cluster 1 (Figure 3). It necessari- jection) between these less massive clus-
background of GRSMC 43.30-0.33 but ly invokes the classical triggering scenario ters and Cluster 1 is ~6 pc, which is larger
further work would have to be done to of Elmegreen & Lada (1977) as one can than the bubble of ionised gas surround-
confirm this. In the third panel we clearly
detect 4 clusterings of reddened sources.
These make up the stellar population of
W49A and some are still visible in the
fourth panel where we find sources
extincted by over 32 magnitudes of visual
extinction, more than half associated with
the newly found clusters. The positions of
these 4 clusters are given in Table 1.

THE PRODUCT OF
THE W49A STARBURST Figure 5: (H-K ) vs. K Colour-
s s
Based on H-Ks colour and Ks magnitude, Magnitude diagram for our survey.
The solid line represents a 1 Myr
we preliminarily identify more than 100 old population taken from the
O-stars candidates associated with the Geneva tracks (Lejeune & Schaerer,
W49A region. We should bear in mind 2001) and the slanted lines repre-
that due to the very high values of fore- sent a reddening of AV = 48 mag.
The black circles identify stars like-
ground and local (inhomogeneous) ly associated with the W49A clus-
extinction we are not complete even for ters. The 90% completeness limit
the most luminous stars in W49A. Never- for a star with errors less than 15%
theless, this number compares already is marked as a bold grey line.

Alves J. & Homeier N., Our Own Starburst © ESO - December 2003 37
 and that it extends over a larger area than
previously thought (see Figure 7 for a
visual comparison with well known young
Galactic clusters). Moreover, star forma-
tion in W49A is still ongoing (the GMC is
not exhausted yet) as 6 hot cores (the pre-
cursors of UC HII regions) were recent-
ly found in the vicinity of the Welch ring
(Wilner et al., 2001). Although W49A is
clearly of the class of 30 Dor in the Large
Magellanic Cloud (LMC), a size compar-
ison with the most luminous cluster in the
Antennae galaxy (Figure 8), makes clear
that the star forming conditions in the
LMC and our Galaxy have to be radically
different. Most important, our results
show that a considerable part of the stel-
Figure 6: SOFI and NACO JHK colour composites of the only compact HII region that is
s
accessible to NACO using natural guide stars. The object in the centre of the compact HII
lar population of W49A is accessible in
region is a newborn ~80 MA star candidate. Through comparison with models (Freyer et al. the 2 µm window. Further characteriza-
2003) we estimate the age of the HII region to be remarkably young: 4 N 104 yr. This is a very tion of the embedded population (via H
young massive star caught in the rare act of passing from the ultracompact to compact HII and K-band spectra and adaptive optics
region stage. techniques) is called for and will surely
provide much needed information on the
starburst phenomenon seen across the
Universe.

ACKNOWLEDGEMENTS
Figure 7: Size We are pleased to acknowledge Miguel
comparison Moreira for discussions and assistance
between W49A with the observations, Robert Simon for
and 1) Orion, 2)
providing molecular line data from the
NGC 3603, and
3) the Arches BU-FCRAO Galactic Ring Survey on
cluster seen in W49 giant molecular cloud and Chris De
the near-infrared Pree for providing radio continuum data
as if they were of the HII regions associated with W49A.
located at the
same distance REFERENCES
and observed Alves, J., & Homeier, N. 2003, ApJ, 589, 45
with the same De Pree, C.G., Mehringer, D.M., Goss, W. M.
instrument (SOFI 1997, ApJ, 482, 307
on the NTT). Elmegreen, B. G., & Lada, C. J. 1977, ApJ, 214,
725
Gwinn, C. R., Moran, J. M., & Reid, M. J. 1992,
ing it, the giant central H II region in Fig- sible to NACO using natural guide stars. ApJ, 393, 149
ure 3. Also, given the short lifetimes of The object in the centre of the compact Homeier, N., & Alves, J. 2004, A&A in prep.
compact regions and the fact that they region is a newborn ~80 MA star candi- Lejeune, T., & Schaerer, D. 2001, A&A, 366, 538
can be found almost over the entire sur- date. We find that the source is unresolved Mezger, P. G., Schraml, J., & Terzian, Y. 1967,
ApJ, 150, 807
veyed region (e.g., the projected distance down to the resolution of the image (600 Shaver, P.A., & Goss, W.M. 1970, AuJPA, 14, 133
between source CC and W49A South is AU at the distance of W49A) and Simon, R., Jackson, J.M., Clemens, D.P., Bania,
~11 pc) suggests a multi-seeded, largely through comparison with models (Freyer T. M., & Heyer, M. H. 2001, ApJ, 551, 747
coeval, star formation episode in the et al. 2003) we estimate the age of the Welch, W. J., Dreher, J. W., Jackson, J. M., Tere-
W49A. region to be 4 N 104 yr. bey, S., & Vogel, S. N. 1987, Science, 238,
1550
In Figure 6 we present preliminary results Our first results suggests that W49A is Wilner, D. J., De Pree, C. G., Welch, W. J., &
of a diffraction limited imaging follow-up indeed the most massive and youngest Goss, W. M. 2001, ApJLet, 550, L81
on the only compact region that is acces- known star forming region in the Galaxy

Figure 8: Size comparison between W49A

and the most luminous cluster in the Antennae
(starburst) galaxy. W49A is approximately an
order of magnitude smaller. The luminosity
comparison is not fair as the wavelengths of
the two images are different (visible and near-
infrared). Also, while the Antennae cluster is
essentially extinction free, W49A is seen
behind a wall of more than 30 mags of dust
extinction.

38 The Messenger 114

 E XTR APL ANAR S TAR F ORMATION
IN NGC 55
We present the first VLT spectra of two compact Hα-emitting objects, located in the halo of the
edge-on spiral galaxy NGC 55. The detection of stellar continuum and the observed emission-
line characteristics indicate that these objects are extraplanar HII-regions. CLOUDY model
simulations establish photoionisation by single OB-type stars as ionisation mechanism and
finally confirm the HII-region character. Hydrodynamical considerations unambiguously restrict
the origin of these regions to the halo. Their creation was most likely triggered by star formation
activity in the disc below. In this picture the gas clouds, out of which the OB stars formed, could
cool and collapse only between two successive bursts of star formation.

R. Tüllmann 1 ,
M. R. Rosa 2 ,
T. Elw er t 1 ,
D. J. Bomans 1 ,
A. M. N. Ferguson 3 ,
R.-J. Dettmar 1
1Astronomisches Institut,
Ruhr-Universität Bochum,
Germany;
2Space Telescope-Euro-

pean Coordinating Facility,

Garching, Germany;
3Max-Planck-Institut für

Astrophysik, Garching,
Germany

Figure 1: Zooming into the centre of

NGC 55. This Hα-image has been
obtained with UT1+FORS1 at the VLT.
It reveals various spectacular features,
such as filaments, colliding shells,
superbubbles, and a newly detected
Wolf-Rayet nebula of surprisingly high
temperature. Note the extraplanar HII-
region (EHR_1) located within the
extended northern SN shell and EHR_2
which sits on top of a huge elongated
(1kpc) filamentary structure.

© ESO - December 2003 39

 extraplanar star formation (ESF). inclination of i = 80°. Moreover, this

D
URING THE LAST DECADES
the formation of massive Although the mechanism that triggers galaxy reveals violent ongoing star for-
stars in galaxies of differ- ESF still remains unclear, especially in mation in the centre and at least two
ent Hubble-Types has view of low gas densities, we have prominent extraplanar HII-region
been studied in great increasing evidence that star formation (EHR) candidates. All these features can
detail. For spirals this process is consid- occurs at rather unusual sites, such as in a be seen very nicely from Figure 1 where
ered to be almost exclusively concentrat- galaxy halo. the VLT Hα-image obtained with
ed to the disc as the efficiency with which In the following, we investigate the UT1+FORS1 is presented.
gas is converted into stars is highest there. possibility of ESF and its triggering A huge curved filament of gas and
On the other hand, the existence of a mechanism by analysing first VLT data of dust, anchored to the disc, is protruding
young and massive OB-star population in compact extraplanar gas clouds with off the image plane, apparently pointing
the halo of the Milky Way galaxy and embedded stellar sources located in the towards R. A.(2000): 00h15m07s and
other spiral galaxies, such as NGC253, is disc-halo interface of the edge-on galaxy Dec.(2000): -39°12(00). Of particular
widely accepted. NGC 55. interest are the two objects marked by
However, as these stars are detected arrows, whose magnifications are shown
at large distances (∼15 kpc) above the EXTRAPLANAR HII-REGIONS in Figure 2.
star forming disc, their creation and ori- There are fairly compact and isolated EHR_1 has a diameter of 17pc and is
gin are rather poorly constrained. In objects visible in Hα-imaging data for located 0.8 kpc above the disc, whereas
order to clarify the origin of this young NGC 55 (Ferguson et al. 1996) which are EHR_2 reveals a projected distance of
stellar component in the halo different located at distances of up to 1.5kpc above 1.5 kpc and spans 22 pc in diameter.
scenarios have been proposed, such as the disc of this galaxy. From a morpholog- If effects along the line-of-sight are
ejection from the disc as a consequence ical point of view, these objects appear negligible, EHR_1 is located within an
of supernova (SN) explosions, ejection very much like a small scale disc HII- expanding oxygen-bright SN shell that
from stellar clusters as a result of gravita- region with embedded clusters of massive was detected over 20 years ago by Gra-
tional encounters, and star formation in stars formed recently. Interestingly, simi- ham & Lawrie (1982). It would be inter-
the halo itself (Keenan 1992, Ferguson lar regions are also discernable in other esting to learn from similar observations
2002). well known edge-on galaxies, such as in other galaxies if these extraplanar
Estimates of the distance a star could NGC891, NGC3628, or NGC5775. In regions predominantly occur at points
travel at a given speed through the halo order to detect these objects, the target where such shells, created by OB stars
during its lifetime reveals that ’in situ’ galaxy should be nearby to ensure suffi- and SNe, intersect and the gas is piled up.
star formation in the halo is the most ciently high spatial resolution and seen At least the compressed gas at
likely scenario for a large fraction of the close to edge-on (large inclination angles R. A.(2000): 00h15m07s and Dec.(2000):
studied sample of halo stars. If proper of i > 70°), as in this case the halo sepa- -39°11(00) is in favour of this idea.
motions of these stars are reasonably low, rates well from the disc. However, the most important imme-
it should be possible with future investi- What makes the barred spiral galaxy diate result is provided by optical multi-
gations to detect their stellar birthplaces, NGC 55, a member of the Sculptor object-spectroscopy (MOS) and concerns
faint gaseous envelopes in the vicinity of Group, an ideal target to study ESF is its the detection of spatially concentrated
these stars, and thus to further strengthen proximity of only ∼1.6Mpc and its high continuum emission, which originates

Figure 2:Blow-up views of EHR_1 In the northern (left) and EHR_2 in the southern halo (right). Both objects show a dense central core with dif-
fuse Hα-emission in their outskirts which is sharply bounded towards the halo.

40 The Messenger 114

Table 1: Element abundances for
the HII-region and the EHRs as cal- Parameter HII-region EHR_1 EHR_2 Solar
culated by NAT for gas tempera- R23 0.88 0.68 0.72 −
tures of 11500K. Values in brackets
were derived with the empirical 12 + log(He/H) 10.94 W 0.02 (10.94) (10.93) 10.84 (10.93) 10.98
R23-calibration. Solar abundances 12 + log(O/H) 8.05 W 0.10 (8.08) 7.77 (7.61) 7.81 (7.68) 8.71
are compiled from the most recent log(N/O) −1.26 W 0.05 (-1.30) −1.50 (-1.54) −1.31 (−1.50) −0.78
data including Christensen-Dals-
gaard (1998) and Grevesse & Sau- log(Ne/O) −0.85 W 0.10 − − −0.71
val (1998). The average metallicity log(S/O) −1.41 W 0.15 −1.94 −1.69 −1.51
<Z/ZA> has been calculated from log(O+/O) −0.731 −0.133 −0.127 −
oxygen abundances as this element
is the most abundant and efficient log(S+/S++) −0.722 − − −
coolant. <Z/ZA> 0.45 0.10 0.10 1.0

within the more diffuse body of the extra- In addition, knowledge of the gas EHRs reveals substantially lower [O/H]
planar objects (Figure 3). phase abundances also helps to distin- abundances of about 10% ZA and thus
The morphology of the continuum guish between different creation mecha- independently also supports the ESF sce-
and the nebular emission-line distribu- nisms of the extraplanar ionised regions. nario.
tion is direct evidence for stellar sources Rather low metallicities compared to the With metal abundances derived this
responsible for the excitation of these disc abundances would indicate that way, we can visualise for the first time the
regions. Correspondingly, the flux-cali- these regions have formed from almost strong differences in the metal content
brated and background-subtracted spec- pristine local halo material. Relatively along the minor axis of this galaxy. From
tra, integrated along their total spatial high abundances would restrict the origin Figure 5 it is obvious that the gas phase of
extent (Figure 4), are very similar to low of the clouds to material processed in oxygen is less abundant in the halo by
excitation HII-regions. star-forming regions of the disc. about a factor of 4.
For EHR_1, continuum emission is We therefore determined the element In order to reach a better coverage of
very weak and can hardly be seen in the abundances of both EHRs and compared the oxygen abundance along the minor
spectrum presented in Figure 4. However, them to those measured in the disc using axis, Figure 5 also plots the metal abun-
it is clearly visible in Figure 3. Figure 4 two independent methods (R23 and the dance of the Diffuse Ionised Gas (DIG).
also shows the spectrum of EHR_2 which nebular abundance tool (NAT), see Tüll- This gas phase is pushed into the halo of
reveals a continuum much more promi- mann et al. 2003 for details). The results a galaxy by multiple SNe where it is visi-
nent than that found in EHR_1. are shown in Table 1. ble as a diffuse extended Hα-emitting
A comparison with CLOUDY model A comparison between the average gaseous layer surrounding the disc. The
simulations reveals that the ionisation metal abundance of the central disc HII- ionisation of the DIG is maintained most
mechanism of these compact objects is region of NGC 55 (45% ZA) and both likely by photoionisation from stars
most likely photoionisation by single OB
stars (O9.5 to B0). Further analysis of
diagnostic diagrams unambiguously con-
firms the HII-region character.

CONSTRAINING THE ORIGIN

OF THE EHRS
The existence of HII-regions in the
halo immediately raises the question
whether these objects originated from
the prominent extraplanar gas of this
galaxy or have just been expelled from
the disc into the halo.
Ejection from the disc can be ruled
out by hydrodynamical considerations
due to the enormous drag, the gas phase
of these regions encountered on its way
out of the disc into the halo. Even a small
amount of interstellar matter located
along the path would lead to a separation
of cloud and embedded stars due to the
enormous difference in impact parame-
ters cloud vs. cloud as compared to stars
vs. cloud. Therefore, we conclude that
these objects must have formed within
the halo.

Figure 3: MOS-frame covering the “blue”

wavelength region from about 3500Å to
5600Å. Both EHR-spectra, reveal only faint
stellar continuum and line emission con-
trary to the central HII-region in the disc.

Tüllmann R. et al., Extraplanar Star Formation in NGC 55 © ESO - December 2003 41

 Figure 4:Integrated spectra for both EHRs. Although their emission-line characteristics is similar to ordinary disc HII-regions, it appears that
EHRs are ionised by a slightly softer radiation field, as implied by a significantly lower flux in the [OIII]λ5007 emission-line.

located in the star forming disk below. As and (b) what triggered the collapse to Since both EHRs are located above
the DIG is no longer involved in star for- actually form those stars? the central part of NGC 55, it appears
mation processes, it is expected to be also The most simple and natural explana- likely, that their formation was triggered
a good tracer of the metal content of the tion is to assume that clustered SNe dur- by star formation activity in the disk
halo gas. The interested reader is referred ing an early burst of star formation eject- below. In this global picture star forma-
to Dettmar (1992) for a comprehensive ed a significant amount of ionised materi- tion in the disk could stimulate as well as
review. al into the halo. terminate the creation of EHRs.
After star formation stopped the Future work will test the ESF sce-
THE GLOBAL PICTURE extraplanar gas had time to cool, collapse, nario for a larger sample of galaxies,
Two important questions directly and form dense molecular clouds. These investigate initial formation conditions
emanate from the ESF hypothesis: (a) molecular gas clouds, out of which EHRs for EHRs, and check if the central stars
how did gas reach the halo in a quantity have formed, can survive and collapse can separate within their lifetime far
to cool, collapse, and form neutral, dense only in the period between two succes- enough from their birthplaces and con-
clouds from which new stars were born, sive bursts of star formation. tribute to the observed stellar halo popu-
lation.

REFERENCES
Christensen-Dalsgaard, J., 1998, Space Sci. Rev.
85, 19
Dettmar, R.-J., 1992, Fundamentals of Cosmic
Figure 5: Oxygen abundance as a Physics, 15, 143
function of |z|, the distance along Ferguson, A. M. N., Wyse, R. F. G., & Gallagher,
the minor axis of NGC 55. The open J. S., 1996, AJ 112, 2567
z [kpc]

symbol represents the averaged Ferguson, A. M. N., 2002, BAAS 200, 3306
oxygen abundance for the disc, Graham, J. A., & Lawrie, D. G., 1982, ApJ 253,
whereas the error bar represents L73
data published by other authors. Grevesse, N., & Sauval, A. J., 1998, Space Sci.
The data-point labelled “HR” has Rev. 85, 161
been slightly shifted along |z| to Keenan, F. P., 1992, QJRAS 33, 325
separate error bars. The one named Tüllmann, R., Rosa, M. R., Elwert, T., Bomans,
“DIG” represents a special compo- D. J., Ferguson, A. M. N., & Dettmar, R.-J.,
nent of the ISM at intermediate |z|- 2003, A&A, 412, 69
distance (see text for details). log (O/H)

42 The Messenger 114

 F L ARES FROM THE DIRECTION OF THE
B L A CK H OLE IN THE G AL A CTIC C ENTRE
N RECENT DRAMATIC OBSERVA-

I TIONS using NAOS-CONICA on

the VLT, near-infrared flares from
the direction of the the black hole
at the Galactic centre have been
detected. The signals, rapidly flickering on
a scale of minutes, must come from hot
gas falling into the black hole, just before
it disappears below the "event horizon"
of the monster. The new observations
strongly suggest that the Galactic Centre
black hole rotates rapidly. Never before
have scientists been able to study phe-
nomena in the immediate neighbourhood
of a black hole in such a detail. These
results were published recently by Rein-
hard Genzel and colleagues* in the jour- Figure 1 shows the detection of one of these flares, obtained with a resolution of 0.040 arcsec
nal Nature**, and were presented in ESO in the near-infrared H-band (λ=1.65 µm). The observations were made with the NACO imager
Press release PR 26/03. on the 8.2-m VLT YEPUN telescope at Paranal on May 9, 2003. The image covers a sky area of
This flare, and several others like it, were about 1 x 1 arcsec, corresponding to about 45 light-days at the distance of the Galactic Cen-
coming from exactly the direction of the tre. The time (in minutes from the beginning of the data set) is shown at the upper right of each
image. North is up and East to the left. The position of the 15-year orbiting star S2 is marked
supermassive black hole at the heart of by a cross and the astrometric location of the black hole is indicated by a circle.
the Milky Way. The team members were
certain that the black hole must be accret-
ing matter from time to time. As this mat- tions of General Relativity. The most also measured its spin for the first time.
ter falls towards the surface of the black striking result is an apparent 17-minute And that turns out to be about half of the
hole, it gets hotter and hotter and starts periodicity in the light curves of two of maximum spin that General Relativity
emitting infrared radiation. But no such the detected flares. If this periodicity is allows. As Reinhard Genzel comments,
infrared radiation had been seen until caused by the motion of gas orbiting the “the era of observational black hole
that night at the VLT. black hole, the inevitable conclusion is physics has truly begun”.
A careful analysis of the new observa- that the black hole must be rotating rap- Variability on time scales of one hour to
tional data, reported in the Nature article, idly. several days was also observed in late
has revealed that the infrared emission It is known from theory that a black hole May/June 2003 at 3.8 microns at the Keck
originates from within a few thousandths can only have mass, spin and electrical telescope by a team of observers lead by
of an arcsecond from the position of the charge. Last year the team was able to Andrea Ghez (UCLA, paper in press ApJ
black hole (corresponding to a distance of unambiguously prove the existence and Letters). As in the case of the VLT data,
a few light-hours) and that it varies on determine the mass of the Galactic Cen- the Keck observers find the variable L-
time scales of minutes. This proves that tre black hole (ESO Press Release 17/02). band source to be coincident with the
the infrared signals must come from just If their assumption is correct that the black hole position to within less than 10
outside the so-called "event horizon" of periodicity is the fundamental orbital mas.
the black hole, that is the "surface of no time of the accreting gas, they now have (based on ESO Press Release 26/03)
return" from which even light cannot
escape. The most likely emission process
of the infrared emission is synchrotron
emission from relativistic electrons. The
observed intensity of the infrared emis- Figure 2 displays the "light curve" of a
sion and its spectral energy distribution light flare from the Galactic Centre, as
suggest that a significant fraction of the observed in the K-band (λ=2.2 µm) on
electrons near the event horizon are June 16, 2003. This and a second flare dis-
accelerated to energies much above the covered about 24 hours earlier show vari-
ability on a time scale of a few minutes and
virial equilibrium, in a non-thermal distri- appear to show larger variations (arrows)
bution, perhaps through magnetic recon- with a 17-minute periodicity. The rapid vari-
nection events as in solar flares. The rapid ability implies that the infrared emission
variability seen in all data obtained by the comes from just outside (the event horizon
VLT clearly indicates that the region of) the black hole. If the periodicity is a fun-
around the event horizon must have damental property of the motion of gas
chaotic properties. orbiting the black hole, the Galactic Centre
black hole must rotate with about half the
The team members have commented that maximum spin rate allowed by General Rel-
the data give unprecedented information ativity. The present observations thus probe
about what happens just outside the the space-time structure in the immediate
event horizon and let us test the predic- vicinity of that event horizon.

* The team consists of Reinhard Genzel, Rainer Schödel, Thomas Ott and Bernd Aschenbach (Max-Planck-Institut für extraterrestrische Physik, Garching, Germany), Andreas
Eckart (Physikalisches Institut, Universität zu Köln, Germany), Tal Alexander (The Weizmann Institute of Science, Rehovot, Israel), François Lacombe and Daniel Rouan (LESIA
-Observatoire de Paris-Meudon, France).
**R. Genzel et al. (2003) Nature 425, pp. 934-937: "Near-IR Flares from Accreting Gas around the Supermassive Black Hole in the Galactic Centre"

© ESO - December 2003 43

 R OSES IN THE S OU THERN S KY
T HE W IDE -F IELD -I MA GER AT L A S ILL A U NVEILS I NTRIC ATE
S TRUCTURES I LLUMINATED BY H OT S TARS
(ESO PR PHOTOS 31a-e/03)

44 IN THE LARGE MAGELLANIC CLOUD is a spec- tributed by emission from singly-ionised oxygen atoms (shining

N tacular example of a giant HII region. Having

observed it in 1999 (see ESO PR Photos 26a-
d/99), a team composed of Fernando Comerón
and Nausicaa Delmotte from ESO, and Annie
Laval from the Observatoire de Marseille (France), again used
the Wide-Field-Imager (WFI) at the MPG/ESO 2.2-m telescope
of the La Silla Observatory, pointing this 67-million pixel digital
at the ultraviolet wavelength 372.7 nm), while the green colour
comes from doubly-ionised oxygen atoms (wavelength 500.7
nm). The red colour is due to the Hα line of hydrogen (wave-
length 656.2 nm), emitted when protons and electrons combine
to form hydrogen atoms. The red colour therefore traces the
extremely complex distribution of ionised hydrogen within the
nebulae while the difference between the blue and the green
camera to the same sky region in order to provide another strik- colour indicates regions of different temperatures: the hotter the
ing - and scientifically extremely rich - image of this complex of gas, the more doubly-ionised oxygen it contains and, hence, the
nebulae. With a size of roughly 1,000 light-years, the peculiar greener the colour is.
shape of N44 clearly outlines a ring that includes a bright stellar The composite photo produced in this way approximates the
association of about 40 very luminous and bluish stars. real colours of the nebula. Most of the region appears with a
These stars are the origin of powerful “stellar winds” that pinkish colour (a mixture of blue and red) since, under the nor-
blow away the surrounding gas, piling it up and creating gigantic mal temperature conditions that characterize most of this HII
interstellar bubbles. Such massive stars end their lives as explod- region, the red light emitted in the Hα line and the blue light
ing supernovae that expel their outer layers at high speeds, typ- emitted in the line of singly-ionised oxygen are more intense
ically about 10,000 km/sec. than that emitted in the line of the doubly-ionised oxygen
It is quite likely that some supernovae have already explod- (green).
ed in N44 during the past few million years, thereby “sweeping However, some regions stand out because of their distinctly
away” the surrounding gas. Smaller bubbles, filaments, bright greener shade and their high brightness. Each of these regions
knots, and other structures in the gas together testify to the contains at least one extremely hot star with a temperature
extremely complex structures in this region, kept in continuous somewhere between 30,000 and 70,000 degrees. Its intense
motion by the fast outflows from the most massive stars in the ultraviolet radiation heats the surrounding gas to a higher tem-
area. perature, whereby more oxygen atoms are doubly ionised and
the emission of green light is correspondingly stronger, cf. Fig. 2.
THE NEW WFI IMAGE OF N44
The colours reproduced in the new image of N44, shown in Fig. More information, including technical information on the
1 (with smaller fields in more detail in Fig. 2−4) sample three images, can be found at http://www.eso.org/outreach/press-rel/pr-
strong spectral emission lines. The blue colour is mainly con- 2003/phot-31-03.html.

Figure 1 (right) shows the southern part of the spectacular N44 region in the Large Magellanic Cloud. The green colour indicates areas that
are particularly hot. The field measures 27.5 u 26.5 arcmin2. North is up and East is left. Figure 2 (above, middle) shows the ionised region
DEM L 159 and two clusters with hot stars named KMHK 840 (top left) and KMHK 831 (bottom right). Figure 3 (above, left) shows a region
with pink-green shades that has been designated DEM L 144 and is located at the bottom centre of figure 1. It is a region of ionised hydrogen.
Note that in figures 2 and 3, the colours have been enhanced compared to figure 1 to clearly show the different shades.Figure 4 (above, right)
shows a part of the central nebula, known as N44C. The green colour indicates areas that are particularly hot. The nature of the exciting source
that delivers the necessary energy has been the subject of studies during two decades but is still not known with certainty.

44 The Messenger 114

 M ETRICS TO M EASURE ESO’ S
S CIENTIFIC S UCCESS
BASED ON PUBLICATION STATISTICS, WE PROVIDE A FIRST ASSESSMENT OF ESO'S SCIENTIFIC IMPACT AFTER
FOUR YEARS OF VLT OPERATION. A BRIEF DISCUSSION OF THE COMPLEXITY OF MEASURING SCIENTIFIC SUCCESS
AND ITS INHERENT PROBLEMS IS GIVEN. WE PRESENT PUBLICATION AND CITATION STATISTICS DRAWN FROM THE
ESO PUBLICATION DATABASE AND PROVIDE SOME PRELIMINARY INTERPRETATION.

B. L EIBUNDGU T, U. G ROTHKOPF, A. T REUMANN (ESO)

SO’S ULTIMATE GOAL is to advance astronomical reproducible by other organisations, and so far no general

E knowledge and provide tools for progress in our

understanding of the world we live in. There is a
long chain of actions, interactions and activities
which lead to such results, and it is important for
organisations to evaluate the impact they have on their specific
research field and on society as a whole.
The scientific procedure is to produce a hypothesis (often
scheme has been developed. One reason was that up to a few
years ago, no uniform non-commercial database was available to
the astronomical community. This has changed with the avail-
ability of the ADS system that collects publications and citations
of basically all astronomical literature. Astronomy is privileged
in that this nearly complete (and free of charge) database exists,
as it represents a situation completely different from almost all
based on previous data and well-known and -loved paradigms), other sciences.
design a test for this hypothesis, carry out the test, analyse the In the following, we present publication and citation statis-
outcome and finally publish the results. It is this last bit, the pub- tics drawn from the ESO publication database with some pre-
lication and dissemination that closes the loop and enables oth- liminary interpretations. After describing how the information is
ers to continue the line of thought. assembled (section 2), we will discuss different criteria that
As an organisation providing resources to the European could serve for interpretation (section 3). In section 4 we pres-
astronomical community, ESO has to evaluate how successful ent the ESO statistics.
and competitive it is compared to other observatories and astro-
nomical institutions. This can be done in many ways, ranging ASSEMBLING THE INFORMATION
from assessing user satisfaction, efficiency of observations and Publications resulting from ESO data should be clearly identi-
quality of delivered data to the achieved scientific break- fied as such. The Call for Proposals specifies that papers must list
throughs. While the first three metrics lie within ESO’s purview, the observing programme(s) within which the data were
the last one is a joint effort of the research community and ESO. obtained in a footnote. This serves several purposes; most impor-
The recent questionnaire on user satisfaction with the service tantly, it helps to measure the observatory’s scientific success.
mode offered by ESO (Comerón et al. 2003, Messenger, no. 113, Astronomers are increasingly following this requirement,
32) and the evaluations given to the Users Committee in the end although not yet as consistently as necessary.
of run reports for visitor mode observations measure how well At ESO, the librarians search all major astronomy journals
ESO is doing compared to the expectations of its users. The sta- for publications deriving from ESO data. When ESO-related
tistics on telescope downtime indicate how well the observato- information is not obvious from the publication, a cross-match
ries function. However, the final aim of the combined work of with the observing schedule is made. Should the records still be
astronomers and ESO is scientific progress. incomplete we contact the first author or PI in order to obtain
missing details. For completeness, the ADS database is queried
MEASURING SCIENTIFIC SUCCESS at regular intervals.
The definition of scientific success is not easy. This becomes This procedure leads to the most comprehensive possible
obvious when one does a small literature survey of previous record of refereed publications based on observations with ESO
studies on scientific success. Comparisons among observatories facilities. A recent comparison with automated searches in ADS
(e.g. Trimble, 1995, PASP, 107, 977; Bergeron and Grothkopf, showed that a considerable fraction of papers is not identified as
1999, Messenger, no. 96, 28; Benn and Sanchez, 2001, PASP, 113, based on ESO data in the ADS database, while others are
385), astronomical institutions (Abt, 1994, PASP, 106, 107) and wrongly attributed to ESO (Grothkopf and Treumann, 2003,
individual astronomers (Burstein, 2000, BAAS, 32, 917) are all LISA IV proc., www.eso.org/libraries/lisa4/Grothkopf1.pdf). The
available. Observatories in particular attempt to quantify their main reason is that automated retrieval tools are not capable of
impact (Meylan, Madrid and Macchetto, 2003, STScI Newsletter, interpreting the context in which search terms appear and thus
20, no.2, 1). Another recent example is the statistical study on cannot discriminate between relevant and irrelevant papers.
the productivity of ESO’s La Silla observatory (see Annex I of The ESO publication database is publicly available through
the La Silla 2006+ report; available at www.eso.org/gen-fac/com- the libraries webpage at www.eso.org/libraries/. Entries contain
mit/ls2006/). Investigations on publication behaviour and pat- authors and title of papers, publication year, journal, volume and
terns have also been published (Abt and Zhou, 1996, PASP, 108, pages as well as ESO-specific information, such as programme
375; Abt, 1998, PASP, 110, 210). IDs, telescopes and instruments used and the observing mode
Most of the studies apply bibliometrics, i.e., the quantitative (service or visitor) in which the observations were carried out.
analysis of average or total numbers of publications and cita- The records are linked to ADS for download of citation infor-
tions. However, they all use specific methods which are often not mation or access to the online version of papers. This database

46 The Messenger 114

was used to assess how ESO’s new and attention, and they have to be put into these publications are rare as the selec-
existing observing facilities impact the context. tion is so severe.
progress in our science. Citations critically depend on how For an observatory it cannot be the
much time has elapsed since a paper was goal to produce high-impact papers itself,
CRITERIA FOR THE INTERPRETATION published. Publications have a mean life but it must aspire to provide the facilities
There are many ways to look at the time after which they are either forgotten which allow astronomers to explore new
impact facilities and scientific collabora- or superseded by more recent findings. territory and make fundamental discover-
tions have. The two extremes are counting Only very few papers become ‘classics’ or ies.
papers and counting Nobel prizes. The even enter textbooks; many will be out- In bibliometrics, typically citation
latter is not meant as a joke, although of dated within a rather short time. The counts determine which publications are
course it is reaching high. It does, howev- mean attention span for papers has short- regarded as ‘high impact papers’. Such sta-
er, illustrate that the interpretation ened from six years in the 1990s (Abt, tistics have been used to compare various
strongly depends on the weights assigned 1996, PASP, 108, 1059) to about two years observatories against each other as well
to different criteria. Reaching too high today (Crabtree and Bryson, 2001, as to argue that certain telescopes are not
will decrease statistics to small numbers JRASC, 95, 259). Electronic publishing ‘competitive’ any longer. One should be
and diminish their meaningfulness. On and a general acceleration in information careful in using such number statistics
the other hand, simple quantitative statis- exchange have contributed substantially blindly as important information may be
tics, ignoring any quality issues, are just as to this decrease. Accordingly, there can be easily overlooked. An example is the
dangerous. It is essential to find the right strong fluctuations in citation statistics ongoing discussion between 4m- and 8m-
balance and to refrain from over-inter- within a relatively short time for recent telescope science. The extra-solar planets
preting results. papers, while older publications that have were discovered at small telescopes (1m
passed their peak will not be cited much to 2m diameter) with the important fac-
Papers longer. tors of time baseline and progress in
Once the data have left the observatory, it A frequently discussed issue are self- analysis software.
is up to the astronomers to convert the citations. While self-cites can influence The evaluation of papers has to take
bits into knowledge. The most tangible statistics for individual papers, their num- into account other phenomena as well. If
results are scientific publications. The ber typically is too low to cause much a publication ‘finishes off’ a field and the
observatory influences the process by concern for whole observatories. Scien- trend moves to other problems, it will not
providing data products that raise the tists simply can not publish fast enough to be cited very often. Many people will
researchers’ interest in writing papers. increase the number of citations of their have personal lists of ‘most important
This requires adherence to strict quality own papers considerably. contributions’ that are not necessarily
standards, in particular when the user borne out in the statistics. Such contribu-
base is growing. Still, some projects may High impact tions are of fundamental importance, yet
not result in any publication for a variety Some publications become highly visible they do not produce the reaction one
of reasons and the goal must be to keep and important for the progress of astron- would expect.
this number as low as possible. The publi- omy. They may introduce a paradigm
cation rate of course depends on the effi- change, produce the fundamental data set ESO STATISTICS
ciency and size of the observatory. An for a given subfield or are seminal The ESO statistics presented in this sec-
observatory that serves a large user com- reviews. Everybody will be able to identi- tion were compiled and analysed with the
munity will generate more papers. fy the five most important papers in their above remarks in mind. They provide a
An increasing number of publications field of research; they enter the syllabus snapshot of the current status (beginning
are based on data from more than one tel- of discussions and are the pillars upon of October 2003) and give a first assess-
escope, and even more than one observa- which research fields stand. By definition ment of the impact of the VLT on astron-
tory. These papers will be regarded by
several organizations as originating from
‘their’ data and hence will appear in sev-
eral statistics.

Citations
The number of citations indicates how
well a publication is accepted and how
important it is considered within the com-
munity. Similar to observatories that do
not produce data suitable for publication,
scientific papers that are not cited are
useless. But simply counting citations is
problematic, and one must be aware of
the pitfalls. Some papers create strong
reactions because they are wrong or mis-
interpret data. Catalogues can generate
many citations as they are the basis for
derivative results. Papers in a popular
field will generally attract more attention
than those which address very specific Figure 1: Number of papers based on ESO data published per year. The data for La Silla (dark
blue) and Paranal (light blue) are shown individually (left scale). In addition, the total number is
topics within a small community. Hence, shown as the red line (right scale). The 2003 values were extrapolated from the first nine
citation statistics are not necessarily a months. Note that papers can be based on data from both observatories, hence the total
measurement of quality, but certainly of number is smaller than the sum of papers from individual sites.

© ESO - December 2003 47

 Figure 2: A steady increase in the numbers of
papers per year is observed for all VLT instru-
ments. The 2003 values were extrapolated from
the first nine months.

omy. Some earlier statistics, in particular special as most publications for this ferent publication behaviour as the way
those on La Silla, have been presented in instrument currently come from commis- astronomers deal with data may depend
the La Silla 2006+ report. We will concen- sioning data. The next years will show on how they were obtained. Also, the two
trate here on the early years of the VLT. whether more complicated instruments modes offer different astronomical
produce similar (or higher) publication opportunities, for instance monitoring
rates. Other important issues could be the projects in service mode. In Figure 3 a
Papers average run length and the larger observ- first comparison of the numbers of papers
The La Silla 2006+ report showed that ing time overheads for IR instruments. derived from service and visitor mode
around 400 publications based on ESO It is interesting to compare these sta- observations is made. The distribution is
telescopes were published each year. tistics to trends observed at other facili- fairly even between the two modes. The
Since 1999, the share of papers coming ties. Numbers are available for some fluctuations are at this point probably sta-
from La Silla telescopes has been fairly ground-based observatories as well as tistical. One has to take into account the
stable; the VLT has generated a steady space missions, like ISO and HST. The delay between observations and their cor-
increase. The number of publications rate of papers continuously increases dur- responding publications so that the origi-
depends of course on how many facilities ing the first few years (see Meylan, nal distribution of modes for the observ-
are offered. While La Silla has reduced Madrid and Macchetto 2003). The com- ing programmes per observing semester
the number of telescopes by closing the parisons are not straightforward, howev- is blurred by the time of publication.
smaller ones, Paranal has seen all four er, as the space missions typically have a Hence, it is too early to draw firm conclu-
VLT unit telescopes come into operation smaller instrument suite and different sions. However, both modes appear
between 1999 and 2002. The number of observing patterns. Ground-based tele- rather successful in producing results suit-
instruments increased, and so did the scopes are mostly confined to night-time able for publication. Among other things
observing opportunities and the fraction observing, while this is not necessarily the this means that the quality of service
of the astronomical community that could case for space missions, where other con- observations matches that from visitor
be attracted. Figure 1 shows the publica- straints play a role. mode runs, and service mode is accepted
tions in refereed journals per year sepa- We investigated whether VLT service as a viable option by the observer com-
rately for the two observatories as well as and visitor mode observations lead to dif- munity.
the total number for all ESO-based
papers. During the first nine months of
2003 already more VLT papers have been
published than in the entire year 2002.
Extrapolating from the first three quar-
ters to the end of the year yields a total
number near 500 publications for this
year with about 240 – or nearly half – of
the papers coming from the VLT. This is a
higher number of papers based on ESO
data than ever before. Overall, it repre-
sents an increase in publications of about
20%.
The increase of VLT papers is a good
sign. We expect this trend to continue in
the near future as more instruments are
added to the observatory.
Figure 2 shows the statistics for the
individual instruments. The increase of
publications per year is comparable for Comparison of publication rates from VLT service and visitor mode pro-
Figure 3:

most instruments. NACO is somewhat grammes. The 2003 values were extrapolated from the first nine months.

48 The Messenger 114

Citations
The citation statistics of ESO papers were
gathered with information provided by
ADS. Although there may be inaccuracies
at the individual level, we believe that
these are mostly negligible for compar-
isons at the scale of observatories. There
appear to be no known systematics in the
ADS system.
It is obvious that citation statistics
depend on the time when they are assem-
bled. The dynamics of citation rates are
beyond this investigation, but one needs
to keep in mind that statistics of recently
published papers can change quickly. This
is illustrated in Figure 4. The citations for
papers published in 2001 nearly doubled
during the first nine months of 2003! This
trend may continue for papers published
in 2002. After a few years, citation rates
reach a peak; afterwards they drop. The Figure 4: Changes of citations from December 2002 to October 2003.
VLT publications have not yet reached
this level, but the early papers are not
cited as much as the more recent ones.
The number of citations per paper is publications typically result in a bifurcat- al. also presented the mean number of
rather high; on average, VLT papers ed distribution. Some papers receive only citations per year for all refereed astro-
receive more than 10 cites after a few few citations even after several years, oth- physics papers and found that publica-
years. As of October 2003, citation rates ers increase their total citations over a tions based on HST data on average are
for papers published in 2001 and 2002 are long time. About 5 to 10% of papers cited twice as often. With a mean citation
significantly higher than those for the based on VLT data achieve rather high rate of approx. 15 for papers published in
very first papers. citation rates (more than 10 citations per 1999, 2000 and 2001, we recognize a simi-
Potentially, there are various mecha- year). More specifically, over 20% of the lar trend for VLT papers. A more detailed
nisms at work. The VLT has received papers published in 2001 show citation analysis is only feasible though with a
greater attention with time. Also, more rates above 10 per year, and 8% received larger baseline.
extensive projects, e.g. Large Programmes more than 20 citations per year two years
that lead to a larger body of data take after publication. While the first two years CONCLUSION
longer to complete. Possibly, papers of VLT operations have not yielded pub- After four years of VLT operation, we
appearing in 2001 and 2002 presented lications attracting more than 50 citations start to see clear signs of the impact the
more comprehensive studies, while the in total, there are six papers from 2001 observatory has on the astronomical com-
early papers may have mostly been letters that exceeded 50 citations within less than munity. The overall publication rate of
and short communications which were two years of publication. One paper from papers based on ESO data − approxi-
superseded by the more substantial 2002 has reached this number within one mately one refereed paper published per
papers following a bit later. Other possi- year of publication. calendar day – remained constant for the
bilities are that with more users and more past few years. In 2003, the total number
papers, the rate of self-citations is going COMPARISON WITH of publications is now increasing, mostly
up with time. This is, however, unlikely to OTHER OBSERVATORIES in line with the growing number of facili-
cause such high citation numbers. At present, the different observatories ties offered at Paranal. La Silla still main-
In general, the VLT seems to produce and organisations assemble publication tains a high publication rate despite a
a healthy reaction from the astronomical statistics according to their respective reduction in available telescopes. With
community and its contributions are policies. As the selection criteria can vary regard to observing modes, no clear pref-
recognised. vastly, comparisons have to be done with erence can be stated. Up to now, service
great care. Meaningful results can only be and visitor mode programmes result in
High impact achieved when statistics are compiled roughly the same number of publications,
As there is no clear definition of what based on the same rules and methodolo- which generally corresponds to the time
high-impact publications are, we restrict gy. This of course requires a close collabo- allocation. The large percentage of highly
this section to a discussion of some ideas. ration among observatories and empha- cited papers and the overall citation sta-
We are able to identify the most frequent- sizes their inter-dependence rather than tistics prove that the scientific results pro-
ly cited papers coming from ESO tele- their competition. duced by the VLT are highly visible and
scopes. This is the first step in such an Bearing these comments in mind, we note well recognised within the astronomical
analysis. Where to draw the line between that the first four years of VLT publica- community.
average publications and those that tion statistics are similar to those of HST Once the number of years over which we
change the way we look at the universe is (see the recently published statistics in look back is sufficiently large to average
probably somewhat arbitrary. In addition, Meylan, Madrid and Macchetto, 2003, out misleading short-term effects, we will
the perception of the community changes STScI Newsletter, 20, no. 2, 1). Both VLT be able to re-investigate the scientific
over time and the impact of publications and HST experienced comparable start- impact of the VLT in a more comprehen-
becomes obvious only in retrospect. up phases with an annual increase in pub- sive study. The results will be published in
We observe that citations for ESO lished papers of approx. 75%. Meylan et a future Messenger article.

Leibundgut B., Grothkopf U., & Treumann A., Metrics to Measure ESO’s Scientific Success © ESO - December 2003 49
Other Astronomical News
R EPOR T ON THE TOPIC AL M EETING

R ESOLVED S TELL AR P OPUL ATIONS

G EORGE H A U & D ANIELLE A LLOIN (ESO)
DVANCES in ground- and space-based observational tion which may be evidence of extra-tidal stars. Like Carina, E.

A capabilities, as well as modelling tools, have enabled

the study of resolved stellar populations with unprece-
dented details. This topical meeting, held at ESO Vitacura on
Hardy showed that Fornax also has a complex star formation
history. The Ca II triplet equivalent widths show substantial
metallicity enrichment, which suggests that metal ejection
June 27, 2003, was aimed at bringing astronomers together to effects must be small. D. Faria described the study of metallicity
examine the progress that has been made in this field. distribution using Strömgren photometry, and I. Saviane
D. Minniti kicked off the Topical Meeting with an introduction described the Relative-Ages project.
to the Local Group and outlined some of the outstanding ques- Are there more Local Group galaxies to be discovered? Prob-
tions. ably very few, said A. Whiting, who is conducting an all-sky sur-
Concerning MACHOS, A. Rest described how the shape of vey.
the LMC luminosity function affects the number of microlensing Going further afield, an exciting development is the possibili-
events expected. ty to resolve the stellar populations in nearby groups. D. Minni-
Tidal streams and mergers were the subjects of several talks. ti described the work on Centaurus A, which has substantial
Deep wide field searches conducted around M31 and M33 show intermediate age stars amongst a metal-rich halo population.
contrasting results. A. Ferguson found a giant stellar stream in More than 1000 Mira long-period variables have been detected,
the outer parts of M31, while M33 shows no evidence of sub- which are used to derive a distance of 4.2 Mpc for this galaxy.
structure. D. Geisler found that the halo of M33 may host an Altogether this Topical Meeting has been very fruitful. Many
intermediate-age population. of the results discussed at the meeting demonstrate the capabil-
Closer to home, G. Marconi finds that the Sagittarius Dwarf ities of large telescopes such as the VLT, and of wide-field
has a metal rich young population with zero alpha-element over- imagers. It is evident that there is a large community of
abundance, possibly resulting from some recent starburst trig- astronomers in Chile working on different aspects of the same
gered by a passage through the Galactic disc. topic. Exchanging information about their programs and results
Complex star formation histories can also be seen in other has been an enlightening experience.
Local Group dwarfs. A. Walker showed that Carina has a well
defined blue-plume of young Main Sequence stars and a narrow Warm thanks go to A. Lagarini for taking care of the logistics
RGB. The star count exhibits a “shoulder” in the northeast direc- of the meeting.

ESO-ESA M EETING ON A C OORDINATED A PPRO A CH TO A S TRONOMY

C OSMOLOGY AND F UND AMENTAL P HYSICS

P ETER S HAVER (ESO)
ASTRONOMERS OF ESA AND ESO SHARE THEIR VIEWS ON FUTURE COOPERATION BETWEEN SPACE AND
GROUND-BASED ASTRONOMY.

between representa- both organizations, followed by discus- practical case discussed at the meeting

A
MEETING
tives of the ESA and ESO sci- sions of scientific areas of common inter- was the joint use of Newton-XMM and
ence advisory structures took est. Areas of overlap and complementari- VLT.
place in Garching (Germany) on 15−16 ty were identified, and possibilities for 4. Initially three working groups are
September 2003, to explore possibilities future coordination were discussed. being established, to consider coordina-
for future coordination between ground A number of actions are being under- tion in the areas of extrasolar planets, the
and space astronomy. Members of the taken as a result of the meeting. scientific exploitation of the Herschel and
executives of both ESO and the ESA 1. A document is being jointly pre- ALMA projects, and the monitoring and
Science Programme also participated. pared on the relevant activities carried study of NEOs.
The meeting took place in the context out by both organizations and the syner-
of the long range planning activities cur- gies identified for future planning. This At the meeting, ESO and ESA recon-
rently underway at both ESO and ESA. information will be widely distributed to firmed their support to the further devel-
The objective was to exchange ideas for the scientific community in Europe. opment of the AVO activities in integrat-
further coordination of activities, given 2. A framework is being outlined on ing access to data archives from ground
the European nature of the two organiza- how to deal jointly with large pro- and space-based facilities. The documents
tions and their service to essentially the grammes requiring space and ground- and reports from the above activities will
same scientific community, and to serve as based support. Further discussion will be released as they become available.
input for the long range studies being car- then lead to a proposal to ESA and ESO
ried out by both organizations. for possible implementation. The agenda and the contributions are
The meeting started with brief 3. A proposal is being prepared for available on http://www.eso.org/gen-fac/
overviews of the current programmes of exchanging time on existing facilities. A meetings/esaeso-2003/agenda.html

50 The Messenger 114

 ESO WORKSHOP ON
“S CIENCE WITH A D APTIVE O PTICS ”
WOLFG ANG B R ANDNER ( MPIA HEIDELBERG) AND M ARKUS K ASPER ( ESO GARCHING)
SEPTEMBER 16 TO 19, 2003, The meeting opened with a brief, yet In recent years, the field of AO has

F
ROM

presented by François Rigaut (Gemini

the ESO Workshop on “Science concise introduction to the history of AO matured considerably, as shown by the
with Adaptive Optics” took place large number of contributions dealing
at ESO Headquarters in Garching. Its Obs.), and continued with overviews of with extragalactic objects. Tim Davidge’s
scope was to bring together users of adap- various AO systems, as well as talks on (Herzberg Institute, NRC) review on
tive optics (AO) from all fields of astron- observing and data analysis strategies. “Resolved Stellar Populations in Star
omy in order to discuss the latest scientif- Christoph Keller (NSO) started off the Clusters and Nearby Galaxies” drew the
ic results obtained with AO systems, and science sessions with a review talk on the link between Galactic and extragalactic
to exchange ideas on how to reduce and results of high spatial resolution observa- studies, followed by several talks on NGC
analyse such observations. tions of the Sun. The session on solar 1068 and other nearby AGNs, and quasar
More than 100 researchers working in astronomy was followed by talks on Solar host galaxies. One of the highlights here

(review by François Menard, LAOG),

many different areas of astronomy came System objects, circumstellar discs was the measurement of the inner rota-
together, providing a comprehensive pic- tion curve around the central black hole
ture of the utilization of AO, and high- substellar companions, HII regions with a in Centarus A as presented by Nadine
lighting the unique science potential of focus on Orion, as presented by Daniel Häring (MPIA). The workshop closed
AO for all branches of astronomy. Rouan (Obs. de Paris) and colleagues, with an outlook on “Science with future
September 2003 also marked the comple- and starburst environments. Hideki Taka- AO systems”, and with a review by Rober-
tion of the first year of science operations mi (SUBARU telescope) presented dif- to Gilmozzi (ESO) on “Science with
of NACO and the first VLT AO System ferential spectroscopy of the extended OWL”.
(Brandner et al. 2002, ESO Messenger 107, 1). molecular layers of late-type stars. The In summary, the 54 talks and more than
Both the “Lessons Learned” by the highlight of the session on “The Galactic 25 poster contributions gave a lively pic-
ESO staff and community, and a signifi- Center and beyond” was the review talk ture of the multitude of science topics to
cant number of science results obtained by Reinhard Genzel (MPE), summarizing be addressed by AO, and proved that AO
with NACO were presented at the work- the latest results on the supermassive has become an essential tool in observa-
shop. black hole in the centre of our Galaxy. tional astronomy.

Technical and ethnological excursions during

the workshop: Visit to the Laser lab at MPE, the
VLT Laser Guide Star PARSEC (above), Rein-
hard Genzel explaining SPIFFI to workshop par-
ticipants (above right), happy workshop partici-
pants visiting the opening of Oktoberfest after
the end of the meeting (right). All images are
courtesy of Dimitrios Gouliermis.

© ESO - December 2003 51

 ESO-WORKSHOP ON

H IGH R ESOLU TION I NFR ARED S PECTROSCOPY

IN A S TRONOMY
H ANS U LRICH K Ä UFL AND A L AN M OOR WOOD (ESO)

at a resolu-

I
NFRARED SPECTROSCOPY
tion of a few km/s allows to study
rotational-vibrational transitions of
many abundant molecules as well as
important atomic lines in a multitude of
interesting astrophysical environments.
The ESO VLT will shortly be equipped
with two unique infrared spectrometers
which combine spectral resolution with
spatial resolutions of ≈ 0.2 arcsec:
N CRIRES, an adaptive optics fed
1−5 µm spectrograph with λ/∆λ≈10 5
N VISIR, including a mode with
λ/∆λ ≈ 3N10 4 between 8−13 µm.
The Workshop held in Garching on
November 18−21, 2003 was organized
mainly to create awareness of these new
observational capabilities and to stimu-
late their use by the community. The first
positive result was that more than 100
astronomers braved the November
weather and made it to Garching, twice as
many as attended a similar workshop in
1992 devoted to both optical and infrared
high resolution spectroscopy. Many late
applicants also, unfortunately, could not Owing to the average Garching November weather the participants had to be squeezed into the
ESO central staircase for a group photo.
be accommodated. The second clear dif-
ference compared to 1992 was the
breadth and maturity of the scientific observational projects, sophisticated fields of stars, studies of ISM chemistry
interest despite the still limited high reso- models and laboratory studies were pre- and the kinematics of stars and gas in
lution instrumentation available. sented covering as diverse fields as our galactic centres.
There were 26 invited, 29 contributed solar system, star-formation and young As the “cement” for this conference
and 39 poster papers covering the observ- stars, “normal” stars, late-type stars, AGB was infrared spectroscopy many people
able universe between 0.2 AU and sever- stars and post-AGB-Objects, the possible came together from different fields who
al Mpc. State-of-the-art instrumentation, direct detection of exoplanets, measure- would normally not meet at conferences
observational highlights, exciting new ments of the abundances and magnetic focused on a specific astronomical topic.
There is thus a good chance that the
workshop has fostered new contacts and
collaborations. Certainly the participants
went away very happy and we at ESO
Intrigued by the aesthetic beauty of learned a lot which is relevant to the fine
this creation of our colleagues in ear- tuning and putting into operation of the
lier epochs, the photo of the Nebra CRIRES and VISIR instruments.
Disc (courtesy of the Landesmuseum
für Vorgeschichte of Saxony-Anhalt, In the next issue of The Messenger
Germany) was selected as the basis there will be an in-depth article summa-
of the workshop poster. rizing the highlights and Lessons Learned
This 32 cm diameter bronze disc, of this particular workshop.
which is by itself already very fasci-
nating, appears even more thrilling
given the fact that this artifact is Those interested in the programme or
being interpreted by archeo- any other details please consult the web-
astronomers as a pocket-version of page: http://www.eso.org/gen-fac/meetings/
the Stonehenge site. For more infor-
mation, please look at the conference ekstasy2003.
web site.

52 The Messenger 114

ESO-USM-MPE WORKSHOP ON

M ULTIWAVELENGTH M APPING OF
G AL AXY F ORMATION AND E VOLU TION
A. R ENZINI (ESO Garching)
O-ORGANIZED with the Univer- observed luminosity density of the ioniz- data (Garth Illingworth). While we still

C sitäts-Sternwarte-München and
the Max-Planck-Institut für
Extraterrestrische Physik, this workshop
ing radiation (from either galaxies or
AGN) does not seem to increase fast
enough with redshift, leaving open the
know very little about the galaxy popula-
tion beyond redshift ∼ 5, these studies
demonstrate that observing strategies
was held in Venice, Italy, from October 13 first two questions above, and highly exist to make rapid progress.
through 16, 2003. The venue, on the prem- speculative our description of the uni- Mostly focused on AGNs, a full session
ises of the Venice International University verse between z ∼ 6.6 and ∼1000. was dedicated to X-ray observations, also
on the island of S. Servolo, proved ideal The search and characterization of the in combination with optical and sub-mm
for hosting the 173 participants for the highest redshift galaxies was the next observations. The space density of both
four full days of the meeting. major topic. Quite successful has been the obscured and unobscured AGNs appears
The workshop was meant to expand narrow-band filter technique to find Ly-α to peak at the fairly low redshift z ∼ 0.7,
over a broader range of issues compared emitters, especially with the Suprime and then declines steadily at higher red-
to the ESO-USM workshop on The Mass Camera on Subaru (Yoshiaki Taniguchi, shifts (Günther Hasinger), thus exhibiting
of Galaxies at Low and High Redshift that Len Cowie), with several objects at z ∼ 5.7 a quite different behaviour compared to
was held in 2001 in the same place. Much and z ∼ 6.6 having been identified thanks the cosmic star formation history, which
indeed has occurred in the meantime, to Ly-α passing through OH-free spectral reaches a maximum at z ~ 2 and then
worth reporting and discussing. An windows. Promising results from the stays nearly constant, as several speakers
impressive set of facilities, in space and on dropout technique were also reported, reported (e.g., David Elbaz, Mauro
the ground, are now used to map galaxy using ACS multicolour data from public Giavalisco, Chuck Steidel).
populations in all relevant windows, from GOODS (Matt Lehnert) and ACS/GTO With much of the star formation being
X-rays to radio wavelengths, from the
local universe to the highest possible red-
shifts, from pencil-beam probes to wide
scale surveys. The primary goal of the
meeting was to document these observa-
tional efforts while trying to answer the
current main questions on galaxy forma-
tion and evolution:
N When did star and galaxy formation
begin?
N What kind of sources have been
responsible for the re-ionization of
the universe?
N How has the overall star formation
evolved with time?
N What is the interplay between galaxy
and AGN formation and evolution?
N What has been the relative role of
thermonuclear vs. accretion power in
the global energetics of the universe?
N How did the mass assembly of
galaxies proceed with cosmic time?
N When did the morphological differen-
tiation of galaxies take place?
N At what pace have these processes
proceeded as a function of the LSS
environment?
Following cosmic time, presentations
and discussions proceeded from high to
low redshift, starting on the first day with
an exciting review of the WMAP major
results by Licia Verde. The most intriguing
result was, of course, the very high red-
shift of re-ionization advocated by the
WMAP team. As then emphasized by The mosaics of the GEMS (blue) and GOODS (magenta) individual ACS exposures in
Figure 1:
Piero Madau and Andrea Ferrara, the the CDF-S field (courtesy of Hans-Walter Rix). ACS data are publicly available.

© ESO - December 2003 53

 bright ones, as one may have expected.
ACS images confirm that the vast majori-
ty of red-sequence galaxies are morpho-
logically early type (i.e., Sersic index
n M 2).
Chuck Steidel reported how he was able
to “colonize” the so-called “redshift
desert” (1.5 m z m 2.5) using LRIS-Blue at
the Keck telescope, hence sampling the
rest-frame UV of star-forming galaxies
selected by a two-colour criterion, similar
to the one first used for LBGs at z ~ 3. Fig.
2 shows the wealth of galaxies that can be
seen in the desert once a UV-blue sensitive
spectrograph is used, as indeed required to
detect the interstellar absorption lines
over the UV continuum of star-forming
galaxies (see an example in Fig. 3).
Compared to surveys of UV/optically
selected galaxies, the observation of
galaxies selected in the near-IR (such as
in the FIRES, MUNICS, GDDS and the
K20 projects) offers a somewhat differ-
ent, complementary view of the distant
universe. Ultradeep near-IR observations
over the HDF-S have revealed a popula-
tion of galaxies with photo-z ~ 3 which
only marginally overlaps with LBGs,
nearly doubling the estimated stellar mass
Figure 2: The redshift distribution of 2-colour selected high redshift galaxies observed at Keck
at this redshift (Ivo Labbé). The growth
with LRIS-Blue and LRIS-Red (courtesy of Chuck Steidel).
with cosmic time of the stellar mass den-
sity was one of the central themes of the
hidden by dust in the optical and near-IR, field, astronomers are eagerly waiting for conference, along with the number densi-
we can learn a lot from mm and sub-mm all the promised complementary data to ty evolution of very massive stars. Both
observations of high redshift galaxies. So, become available. Besides completing the the K20 and the GDDS spectroscopic sur-
while waiting for ALMA, reports of near-IR coverage, this includes especially veys (Andrea Cimatti, Emanuele Daddi,
SCUBA, MAMBO, SIMBA and IRAM the SIRTF mid-IR coverage now to be Adriano Fontana, Hsiao-wen Chen)
observations (respectively by Ian Smail, completed in 2004, along with the FORS2 appears to indicate that much more many
Frank Bertoldi, Tommy Wiklind, and and VIMOS spectroscopy. While the massive galaxies were already in place at
Reinhard Genzel) were followed with GOODS database is still largely incom- zM 1.5 than so far predicted by semi-ana-
much interest. In particular, Reinhard plete, important scientific results are nev- lytical models. Yet, while their number
reported the IRAM measurement of the ertheless being produced with what is density appears to be in better agreement
very high mass of a z=2.8 galaxy via the already in hand. In particular, a robust with a PLE model, they are in fact star-
CO linewidth, showing that the sub-mm estimate of the UV luminosity density burst galaxies much different from the
can do much more besides offering a bet- and of the space density of Lyman-break galaxy population in the model itself. On
ter measure of the star formation rate. As galaxies (LBG) all the way to z ∼ 6 was the other hand, the discrepancy with
SCUBA angular resolution is too coarse, presented (Giavalisco), along with the respect to semi-analytical models seems
in a fully multifrequency approach Scott evolution of their size (Ferguson), while less severe in the MUNICS photo-z sur-
Chapman used high-resolution observa- combining Chandra X-ray and ACS vey (Niv Drory), which is ~ 0.5 magnitude
tions in the radio to identify the likely observations it has been possible to esti- shallower in K but extends over 20 times
optical counterparts of a sizable sample of mate the space density of high-z QSOs in larger area. In the end, everybody agreed
SCUBA sources, then obtained their red- the GOODS fields (Cristiani). that cosmic variance may go some way
shifts using optical emission lines. The The Combo17 project, along with its towards accounting for apparent discrep-
objects lie at a median redshift 2.4, and HST/ACS “GEMS” extension (see Fig. 1) ancies between different surveys, along
with a space density over 1000 times high- was presented by Hans-Walter Rix and with different selection effects.
er than in the local universe they con- Eric Bell. Some 30,000 photometric red- The formation and evolution of early-
tribute significantly to the global star for- shifts from 17 intermediate and broad type galaxies remains a key issue that was
mation rate at their epoch. bands led to an estimate of the evolution widely discussed at the meeting, using
The central part of the workshop was of the red-sequence (or early type) galax- both the low-redshift and high-redshift
occupied by multifrequency surveys. The ies all the way to z ∼ 1, with their colour evidence (e.g., Emanuele Daddi, Daniel
status, first results, and perspectives of change being consistent with passive evo- Thomas, Nobuo Arimoto, Mariangela
GOODS were reported by Mark Dickin- lution. Bernardi, Guinevere Kauffmann, Rober-
son, Mauro Giavalisco, Stefano Cristiani, However, their luminosity density at to de Propris, Scott Trager) with the pre-
and others. With the ACS data fully z ∼ 1 falls short by a factor 2−3 with vailing opinion favouring a very early for-
acquired, reduced, and publicly released respect to a pure luminosity evolution mation epoch (z M 2−3) for the bulk of
along with the VLT/ISAAC near-IR cov- (PLE) model, with the missing galaxies stars in these galaxies, with rather small
erage of about 1/3 of the GOODS-South being the faint ones, rather than the dependence on the environment.

54 The Messenger 114

Figure 3: VLT/FORS2 coadded spectra of star-
burst galaxies with 1.7 m z m 2.3 from the K20
sample (courtesy of Emanuele Daddi), are
compared to the coadded spectra of the 25
bluest and 25 reddest classical z= 3 Lyman-
break galaxies.

While the meeting was mostly dedicat-

ed to observations, hence dominated by
observers, a few theorists also attended
and made lively contributions to it
(Avishai Dekel, Cedric Lacey, Rachel
Somerville, Simon White). Rachel, in par-
ticular, presented efforts in tuning model
parameters trying to push the assembly of
massive galaxies towards earlier epochs.
Imaging/spectroscopic surveys that
have just started were also illustrated,
showing early results from the VLT
VIMOS Deep Survey (VVDS, Olivier Le
Fèvre), the Keck DEEP/DEEP2 survey
(David Koo, Jeffrey Newman), and
GALEX (Chris Martin, Mike Rich). For
other major surveys that are about to
start, motivations, plans, and expectations
were also illustrated, including SWIRE
on SIRTF (Alberto Franceschini, Seb
Oliver), the near-IR Ultra Deep Survey
which is part of UKIDSS (Omar Almaini)
and the COSMOS 2-square degree ACS means they need to cope with cosmic this cursory summary. Much of the suc-
survey (Nick Scoville). Given its conven- variance while mapping galaxy and LSS cess of the meeting was also due to the
ient equatorial/10h location, COSMOS is evolution all the way to at least z ~ 3. 30-minute long discussions at the end of
attracting virtually every major facility on All in all, during the four days of the each session, and to the colloquial atmos-
the ground and in space with the goal of meeting 66 oral and over 100 poster con- phere favoured by the city and by the
providing a full multiwavelength, public tributions were presented, and I apolo- daily vaporetto trips to and from the
dataset, thus promising astronomers the gize for the many I could not mention in island.

T HE VT-20 04 E DUC ATIONAL P ROGR AMME - A U NIQUE O PPOR TUNIT Y

JUNE 8, 2004, Venus starts officially on January 1, 2004, but provisional information is
O N
passes in front of the Sun
as seen from the Earth. This
already available at the dedicated website (www.eso.org/vt-
2004/). When ready, it will provide access to a wealth of related
very rare event (the last one information in many (European) languages about many different
was in 1882 and no living per- aspects (scientific, technical, historical etc.) of this event.
son has ever seen one!) lasts The VT-2004 project invites active participation of all interested
about 6 hours and will be visi- individuals (including teachers, students, amateur astronomers,
ble from most of Europe, etc.) and educational institutions (planetariums, public observa-
Africa and Asia. It will most tories, science centres, etc.). It will provide comprehensive infor-
certainly generate unpre- mation about the related - scientific, technical, social and histor-
cedented attention from the ical - aspects. It encourages and will coordinate real-time meas-
media and the public, not just urements of the transit, thus publically re-enacting the determi-
in these areas, but all over the nation of one of the most fundamental astronomical parameters,
world. the distance from the Earth to the Sun. It also explains the rela-
http://www.eso.org/vt-2004 The VT-2004 project is tion of this event to the search for extra-solar planets by the tran-
launched in this connection sit method, the only one which, in the near future, might be able
and aims at transforming public curiosity into knowledge and to discover Earth-size planets.
interest in science through a broad set of actions. It is managed The VT-2004 project promotes international collaboration
by the European Southern Observatory (ESO) and the European throughout a large part of the world, by observing the same rare
Association for Astronomy Education (EAAE), together with the celestial event, debating it via the web and adding local obser-
Institut de Mécanique Céleste et de Calcul des éphémérides vational contributions to a large, common database. Real-time
(IMCCE) and the Observatoire de Paris in France, as well as the feed-back via the web and the media will ensure that this will
Astronomical Institute of the Academy of Sciences of the Czech become a very special public event. A large, international net-
Republic. The programme is supported by the European Com- work of educational institutions that will be actively involved in
mission in the frame of the European Science Week 2004. It the Venus Transit event is being established (see the website).

© ESO - December 2003 55

 P HYSICS ON S TA GE 3: L IFE AND N EW F RONTIERS
B. Mac k o wiak and R. West (ESO)

TEACHING OF PHYSICS in During the ensuing discussion, the partic-

T
HE
Europe’s schools is changing. In a ipants explained about their experience
growing number of places, thanks with interdisciplinary teaching and made
to well-informed and dedicated teachers, various proposals on how to fuse ALMA
it is becoming an increasingly fascinating and education. They were told to send in
subject, appreciated by the students. their ideas to ESO’s Educational Office,
More and more educators are beginning thereby contributing to the start-up of
to realize that physics lessons may be a this new educational project.
stage for demonstrations of how our daily As before, ESO had set up a stand at
life is influenced by numerous physical the fair, informing the POS 3 participants
phenomena and processes. Moreover, as about this organisation’s main goals, as
new frontiers of research open, new well as its present and future projects, and
ALMA presentation session.
opportunities arise for interesting and leading to interesting and useful discus-
effective teaching means and methods. sions with the teachers. Also at POS 3,
The new trends in the teaching of “experimental” session to demonstrate ESO and EAAE announced the winners
physics and basic facts about this and new and exciting possibilities of the wide of this year’s European student contest
related subjects were high on the agenda and fascinating field of interdisciplinary “Catch a Star!”.
when more than 400 delegates from 22 teaching, with the ALMA project at the Spectacular and original performances
European countries met during this centre and the title “The Atacama Large by students and professional actors, inten-
year’s “Physics on Stage 3” festival (POS Millimeter Array Project and Related sive encounters at the central fair, semi-
3), organized at ESTEC/ESA (Noord- Educational Opportunities”. Emphasis nars and workshops were the components
wijk, The Netherlands) on November 8− was placed on the opportunity to illus- of the rich, one-week POS 3 programme.
15, 2003 by the EIROforum Working trate the workings of a major internation- Among the highlights were the Opening
Group on Outreach and Education, and al science and technology project, not just Ceremony with the attendance of Prince
held with support from the European through research goals and techniques, Johan Friso of the Netherlands and the
Commission under the auspices of the but to introduce other fields, for example Dutch Minister for Education as well as
European Science and Technology Week. geography, chemistry, biology and history. the Farewell Dinner with the presentation
Following the preceding, vastly successful The subject of ALMA can thus be made of the Project Development Awards. Four
events in 2000 and 2002, the main theme more “interesting” and useful in an edu- teachers were awarded cash sums for the
this year was “Physics and life”, reflecting cational context because of the many further development and dissemination
the decision to broaden the Physics on entry points, be it geological aspects (vol- of their excellent projects. A documen-
Stage activities to encompass more of the canoes, earthquakes), historical (the tary film was shown about the solar race
natural sciences, in particular biology, native people in the Atacama region), across Australia and the presentation of
within an interdisciplinary approach. biological (the sparse life in the desert or the Netherlands winner crew proved that
On the first day, the seven EIROforum the high-altitude effects on human physics is life.
organizations (CERN, EFDA, EMBL, beings) or political/international ones The next event in this series will be
ESA, ESO, ESRF, ILL) presented select- (the making of the ALMA project; man- “Science on Stage 4”, to take place in
ed aspects of their current work during agement; operation). About thirty teach- Grenoble (France) in the first half of
individual 3-hour sessions. An eighth ses- ers from more than a dozen countries lis- 2005.
sion was organized by the European tened to talks given by Peter Shaver, Tom Look at the website of the ESO Educa-
Physical Society and the European Asso- Wilson, Bernhard Mackowiak and tional Office (www.eso.org/outreach/eduoff/)
ciation for Astronomy Education Richard West; each of them received a for more information and links to related
(EAAE). ESO had chosen to run an comprehensive booklet and a souvenir. programmes.

Prince Johan Friso (right) of the Netherlands A display during the POS 3 festival. Visualising acoustic waves.
inaugurated the one-week event.

56 The Messenger 114

 F ELLOWS AT ESO
GEORGE HAU NURIA HUELAMO ELENA MASON
I JOINED ESO IN July 2001. Until June I JOINED ESO IN November 2002. I was I ARRIVED AT ESO and Chile in Septem-
2003 I was Instru- supposed to start my Fellowship in ber 2001, as a fel-
ment Scientist for Garching, but decided to move to Chile low with duty in
EFOSC2 on the for a short period of time. I knew the Paranal. I have just
3.6m, whose manu- country because I had visited La Silla started my third
al I have rewritten very often as a PhD student, and I always year of fellowship
recently (see News had the feeling that it was a nice place to which I decided to
From La Silla). live. Professionally speaking I feel I am in spend within ESO
What I like most the right place: Vitacura is a lively insti- too, for the simple
about my work is tute full of special people. reason that I am
the quality of training received. It is very As do many fellows I have my duties in really enjoying
refreshing to arrive at La Silla every time Paranal. From the scientific point of view both the Paranal
to know that I would learn something dif- my experience could not be better: I am and the Vitacura environments. I believe
ferent but invariably interesting. The in touch with sophisticated instruments that the time spent at the observatory,
human factor is also important. I really and it is always fruitful to discuss with vis- despite subtracting working weeks from
enjoyed the interactions with the visiting iting astronomers. Although the work your own research time, provides you
astronomers and to learn about the sci- can be stressful, I find the atmosphere with a competitive knowledge on top
ence they are doing, and to share the amongst the fellows and the rest of instruments, while the contact with col-
excitement when the data arrives! Paranal divisions very good and collabo- leagues and other scientist gives new
Apart from duties on La Silla, I was a rative, which helps to solve the problems ideas and perspectives. Indeed, I am very
member of the Visiting Scientist Commit- in an efficient way. There is also time to happy about this experience.
tee in Santiago. In June 2003 I co-organ- relax and I usually visit the music room
ised a Workshop on Resolved Stellar Pop- with some colleagues. My research focuses on the world of cata-
ulations in Vitacura. clysmic variables (CVs) where I gained
Since July 2003 I am based in Garching Before joining ESO I spent three years in expertise both on short orbital period
with reduced duty. My research is con- Garching where I dwarf novae (DNe), and classical novae
cerned with elliptical galaxy formation at obtained my PhD. (CNe). In dwarf novae I study accretion
different mass scales from a low-redsift I was working in discs and spectral energy distributions,
perspective. Although the trend is to go to the neighbouring continuing the research project started
higher and higher redshifts, I think that a Max-Planck-Insti- during my PhD. While in classical novae I
lot can still be learned from nearby galax- tute (MPE) and am mainly studying their spectroscopic
ies. At the low mass scale I am involved the main topic of evolution. Working on CNe, I have
in a ESO Large Program on the nature of my thesis was the realised the limit of current analyses,
dwarf ellipticals. In the intermediate mass study of magnetic which lack systematic quantitative studies
scale I am comparing galaxy properties in activity and rotation in a group of young and/or modelling for a correct measure-
different environments, such as Kinemat- late-type stars in binary systems. Most of ment of the ejected mass and the shell
ically Decoupled Cores and shells which my thesis was based on X-ray and optical shape. I thus started new projects which
often provide clues on the past merger data. However, I also had the opportunity either make use of new (to me) tech-
history. At the high mass scale I am lead- to work on adaptive optics (AO) data niques (e.g. polarimetry) to investigate
ing a collaboration to study the properties which I found really exciting. I started to the asymmetries in the ejected shell, or
of cD galaxy halos. In several cases we get interested in AO techniques and that intend to recover archive data of CN neb-
found an outwardly rising velocity disper- was one of the reasons why I asked to be ular spectra for a consistent determina-
sion profile which shows that the stars in an ESO fellow: I wanted to work with tion of ejecta abundances. The expertise
the outer parts of cD halos are respond- NAOS-CONICA (NACO). I made the gained on Paranal also gave me the
ing to the gravity of the cluster as a whole. right decision: I am learning something chance to get involved in new projects,
On an unrelated topic, I am also partici- new every day and it is a pleasure to sup- while the lively ESO-Vitacura depart-
pating in an all-sky hunt for Local Group port top scientific research involving this ment offered to me chances for new col-
dwarf galaxies. So far we have found two instrument. As a scientist, I am leading laborations. Both will help me broaden-
dwarfs, but it seems that there isn't a big different scientific projects with NACO ing my knowledge and working field.
population of dwarfs in the Local Group which include imaging and polarimetry of
which have been predicted in some CDM young stars. Moreover, I am collaborating Living in Chile/Santiago one can profit
simulations. with the star-formation people in Vitacu- from the Andes mountain range to do
I think that the fellowship program is ra. This is a very active group with high altitude mountaineering, my major
excellent and I would not hesitate to rec- astronomers working in different hobby during university years. Unfortu-
ommend it to anyone who considers research fields, providing me a wider nately my constantly cracking knee did
applying. It has been fun living in Chile knowledge of different aspects of star for- not help me in a healthy active use of my
and the experience has been fantastic! mation. free time.

© ESO - December 2003 57

Announcements
MARKUS NIELBOCK but then moved to Bochum in order to be Despite the amounts of functional work
MY FIRST CONTACT with ESO was a visit able to concentrate on astronomy. I still at the observatory, there was ample time
at the SEST on La remember a professor in Düsseldorf to pursue my own scientific interests in
Silla in 2000 in claiming that those not having specialised ESO-Vitacura. I am mostly interested in
order to help with on physics in grammar school might not the formation of stars, both in the low-
the installation be suitable for this college career. Well, mass and the high-mass range. In order to
and testing of the this was obviously not true. I found probe the earliest stages of the protostel-
computer facilities physics in school very boring, and the top- lar collapse and thereby look for proto-
for the planned ics I was interested in were not taught stars, I mainly use millimetre and submil-
bolometer array there. So, I helped myself by reading limetre telescopes which made me a bit
SIMBA. This books like Steven Weinberg's “The First exotic among the scientists in Santiago.
matched my scien- Three Minutes”. Already in my PhD thesis I reported on
tific background of early stages of star the detection of numerous low-mass pro-
formation for which I mostly investigate Leaving Europe and going to Santiago tostars and determined their physical
in the millimetre and sub-millimetre, but for a couple of years was a big step and a properties. Recently, I started to concen-
also in the infrared regime of the electro- personal challenge for me, but definitely a trate on the higher end of the mass scale
magnetic spectrum. It was the team rewarding one. What I liked most was the with two projects. The first one deals with
leader Lars-Ake Nyman who encouraged large variety of scientific disciplines the investigation of masers in star form-
me to apply for the ESO Fellowship pro- besides astronomy (cryogenics, micro- ing regions. They are of potential value
gramme. At that time, I was in the middle wave engineering, software programming, for locating young massive protostars.
of working on my PhD in Bochum (Ger- computer system maintenance) I got in The other programme is about the recent
many) under the supervision of Rolf contact with, especially at the SEST, first detection of a massive accreting pro-
Chini, and I really had not thought about where everyone did almost everything. tostar. This is the work of a group of
post-doc positions yet, but the prospect During the more than 2 years (until its German and Austrian astronomers in
sounded very tempting. So, I applied and closure in August 2003), I was also in which I participate. As a next step, we
got the job. But this also meant, I had to charge of the operations and data quality need to constrain the properties of our
speed up and finish the PhD within 2,5 management of the newly installed find more precisely.
years, since the installation of SIMBA bolometer array SIMBA attached to this After more than two years living in Santi-
required me to assume this new position 15-m millimetre telescope. Also the con- ago, I returned to Bochum in September
on June 1st, 2001. tact with colleagues, guests and visiting 2003. But I am looking forward to visiting
I began to study physics in Düsseldorf, observers was very inspiring. Chile for observations next year.

ESO WORKSHOP ON

PLANETARY NEBULAE BEYOND THE MILKY WAY

ESO Headquar
Headquar ters, Garc
Garc hing,
hing, May
May 19-21,
19-21, 20 04

Planetary nebulae (PN) can be detected out to quite large distances The format of the meeting will consist of invited reviews, in the key
from their strong emission lines, principally of [O III]. Photometry of areas of extra-galactic PN research, contributed talks and posters,
the [O III]5007Å emission line allows distance determination through and two discussion sessions.
the Planetary Nebula Luminosity Function. Spectrophotometry of the
emission lines can provide nebular abundances, enabling the use of Central topics will include:
PN as chemical probes of galaxies. The emission lines are also nar- − Surveys for extra-galactic PN
row, making PN ideal kinematic probes of a galaxy's gravitational − The PN luminosity function
potential. Thus the properties of dark-matter haloes can be studied − PN in the Magellanic Clouds
by modelling the PN kinematics. Recently PN have been detected in − Observational dynamics and modelling of PN in galaxies
intra-cluster regions of nearby galaxy clusters. These PN can bring a − Nebular abundance determinations in PN
unique handle on the stars in regions that may harbour a substantial − PN as probes of galactic chemical evolution and
amount of mass. star formation history
− PN in galaxy clusters
Extra-galactic PN serve as versatile probes of nearby galaxies. Plan-
etary nebulae have traditionally been regarded as bright objects;
however in the Local Group, and beyond, they can be faint. Large tel- Scientific Organising Committee:
escopes thus open up the field of extra-galactic PN study. HST and M. Arnaboldi, R. Ciardullo, N. Douglas, K. Freeman, G. Jacoby, R.
Adaptive Optics can provide images of PN in the LMC and SMC, at Mendez, R. Shaw, L. Stanghellini (co-Chair), G. Stasinska, J. Walsh
resolutions previously expected for Milky Way PN. New instrumenta- (co-Chair)
tion is also being exploited to measure PN spectra in bulk. The field
of extra-galactic PN research is developing rapidly and a conference Full details, list of invited speakers and registration information can
to review the progress so far and to chart new developments is now be retrieved from http://www.eso.org/extgalpn04/ or by by email to
timely. [email protected]

This will be the first full workshop dedicated to the subject of extra- Deadline for first registration: 31 January 2004
galactic planetary nebulae. Final deadline: 15 April 2004

58 The Messenger 114

Applications are invited for a position in the Instrumentation Division at the ESO Headquarters in Garching near Munich, Germany.
This post is open to suitably qualified men and women.

Infrar ed A str onomer

Head of the Infrar ed Instr umentation Depar tment
CAREER PATH: V-VI
Assignment: The Instrumentation Division consists of about 30 astronomers, physicists and engineers, who work closely with other Divisions
at ESO plus many national and international Institutes to define, design, build, commission and use cutting-edge optical and infrared instru-
mentation on ESO’s telescopes at its Paranal (VLT) and La Silla observatories in Chile.
You will be responsible and report to the Head of the Instrumentation Division for essentially all aspects of ESO’s infrared instrumentation and
detector programme. This includes follow-up maintenance and upgrading of instruments operating in Chile (SOFI, TIMMI2, ISAAC, CONICA,
VLTI IR detector systems); completion of instruments under development in Europe (VISIR, SINFONI, CRIRES, VLTI and IR detector systems
for Adaptive Optics instruments); design studies of future VLT instruments (HAWK-I, KMOS, ...); maintenance of an active IR detector and con-
troller development programme and definition of future instrumentation for the existing and possible future extremely large telescopes. In addi-
tion to the scientific and technical aspects this will make strong demands on your personal and project management as well as reporting and
presentation skills.
As a senior astronomer you will be a member of the ESO Astronomy Faculty and will be expected to conduct an active research programme.
Education: Ph.D in Astronomy or Physics.
Experience and Knowledge: An outstanding record of both astronomical research and instrument development is required. Excellent commu-
nication skills and command of the English language plus proven team leadership skills are essential. Ideally, you will have experience of man-
aging medium sized projects and teams.
Duty station: Garching near Munich, Germany (with stays at the ESO Observatories in Chile as required e.g. to participate in the commission-
ing of instruments.)
Starting date: As soon as possible.
Contract and Remuneration: We offer an attractive remuneration package including a competitive salary (tax free), comprehensive pension
scheme and medical, educational and other social benefits as well as financial support in relocating your family. The initial contract is for a
period of three years with the possibility of a fixed-term extension or permanence. The title, grade and level of responsibility may be subject
to change according to qualification and experience.
Application: If you are interested in working in a stimulating international research environment and in areas of frontline technology, please
complete the application form to be found at http://www.eso.org/gen-fac/adm/pers/forms/ and send it to Mr R. Block, Head of the Personnel Depart-
ment at ESO together with the names of four individuals willing to provide professional reference letters by 31st January 2004.

For further information, please contact Mr Roland Block, Head of Personnel Department, Tel. +49 89 320 06 589; e-mail: [email protected]. You
are also strongly encouraged to consult the ESO Home Page (http://www.eso.org) for additional information about ESO.

PERSONNEL MOVEMENTS
International Staff
(1 September 2003 - 31 December 2003) FISCHER, Michael (CND), Associate
GIL, Carla (P), Student
ARRIVALS NESVACIL, Nicole (A), Student
EUROPE
ARNDT, Angela (D), Paid Associate DEPARTURES
BERRINGTON, Sylvia (PL), Paid Associate EUROPE
BIALETSKI, Yury (UA), Student ENARD, Daniel (F), Senior Optical Engineer
CAPRONI, Alessandro (I), Associate FRANZA, Francis (F), Opto-Mechanical Technical Engineer
CARMONA GONZALEZ, Andrés (CO), Student GEORGIEVA, Radostina (BG), Associate
CHUZEL, Olivier (F), Application Software Designer/Developer GUIMARAES, Rodney (BR), Associate EIS
CLENET, Yann (F), Associate KOKE, Thomas (D), Administrative Software Systems Specialist
DE BREUCK, Carlos (B), Fellow LIMA, Jorge (P), Associate
DÖLLINGER, Michaela (D), Student NORMAN, Colin (AUS), Associate
EGHOLM, Mathias (DK), Student OLIVIER, Nathalie (F), Associate
ESTEVES, Raul (P), Associate PASQUALI, Anna (I), Astronomer
FIORENTINO, Mauro (I), Science Data Analyst/Programmer WERNER, Daniela (D), Associate
FRANCK, Christoph (D), Opto-Mechanical Engineer CHILE
KÜMMEL, Martin (D), Science Data Analyst/Programmer CUBY, Jean-Gabriel (F), Infrared Instrumentation Specialist
LAING, Robert (GB), European ALMA Instrument Scientist FOUQUE, Pascal (F), Associate Eros II
LARSEN, Soeren (DK), Instrument Scientist RASSIA, Effrosyni (GR), Student
LISKE, Jochen (D), Fellow RÖHRLE, Claudia (D), Student
PEROUX, Céline (F), Fellow SKOLE, Steen (DK), Software Engineer
POTT, Jörg-Uwe (D), Student
RAIMONDO, Gabriella (I), Student Local Staff
RETTURA, Alessandro (I), Student (1 September 2003 – 30 November 2003)
RIELLO, Marco (I), Student
SADIBEKOVA, Tatyana (UZ), Student ARRIVALS
SANTOS, Joana (P), Associate ARRIAS GALANO, Andres, Finance Officer
VAN HEST, Franck (NL), Student EDMUNDS CONCHA, Ann, Executive Bilingual Secretary
WILSON, Thomas (GB), European Project Scientist ALMA MORNHINWEG, Manfred, Execution Engineer, Electronics
WOLD, Margrethe (N), Fellow RITZ SOLARI, André, Procurement Officer ALMA
ZINS, Gérard (F), Associate SALAZAR BARRERA, Daniel, Software Engineer
ZWAAN, Martin (NL), Fellow SCHEN, Tzu-Chiang, Software Engineer
CHILE SEPULVEDA ORTEGA, Jorge, Software Engineer
BAUMONT, Sylvain (F), Student
CHAUVIN, Gael (F), Fellow DEPARTURES
DELLE LUCHE, Céline (F), Student GUTIERREZ CHEETHAM, Pablo, Electronic Engineer
EDEROCLITE, Alessandro (I), Student LOBOS LOBOS, Claudio, Optical Technician

© ESO - December 2003 59

 ESO, the European Southern Observa-
tory, was created in 1962 to “... establish
ESO Workshop Proceedings Still Available
and operate an astronomical observatory
in the southern hemisphere, equipped Many ESO Conference and Workshop Proceedings are still available and may be ordered at
with powerful instruments, with the aim of the European Southern Observatory. Some of the more recent ones are listed below.
furthering and organising collaboration
No. Title Price
in astronomy...” It is supported by ten
countries: Belgium, Denmark, France, 54 Topical Meeting on “Adaptive Optics”, October 2–6, 1995, Garching,
Germany, Italy, the Netherlands, Portugal, Germany. M. Cullum (ed.) 7 40.–
Sweden, Switzerland and the United
Kingdom. ESO operates at three sites in 55 NICMOS and the VLT. A New Era of High Resolution Near Infrared Imaging
the Atacama desert region of Chile. The and Spectroscopy. Pula, Sardinia, Italy, May 26–27, 1998 7 10.–
Very Large Telescope (VLT), is located on 56 ESO/OSA Topical Meeting on “Astronomy with Adaptive Optics – Present
Paranal, a 2,600 m high mountain Results and Future Programs”. Sonthofen, Germany, September 7–11,
approximately 130 km south of Antofa- 1999. D. Bonaccini (ed.) 7 50.–
gasta. The VLT consists of four 8.2 metre 57 Bäckaskog Workshop on “Extremely Large Telescopes”. Bäckaskog,
diameter telescopes. These telescopes Sweden, June 1–2, 1999. T. Andersen, A. Ardeberg, R. Gilmozzi (eds.) 7 30.–
can be used separately, or in combination
as a giant interferometer (VLTI). At La 58. Beyond Conventional Adaptive Optics, Venice, Italy, May 7–10, 2001
Silla, 600 km north of Santiago de Chile E. Vernet, R. Ragazzoni, S. Esposito, N. Hubin (eds.) 7 50.–
at 2,400 m altitude, ESO operates sever-
al optical telescopes with diameters up to
3.6 m. The third site is the 5,000 m high
Llano de Chajnantor, near San Pedro de
Atacama. Here a new submillimetre tele-
scope (APEX) is being completed, and a
Contents
large submillimetre-wave array of 64 TELESCOPES AND INSTRUMENTATION
antennas (ALMA) is under development.
Over 1300 proposals are made each year ALMA Groundbreaking Ceremony . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
for the use of the ESO telescopes. The A. MOORWOOD CRIRES Takes Shape . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
ESO headquarters are located in Garch- R.N. WILSON Addendum to “The History and Development of the ESO Active
ing, near Munich, Germany. This is the
scientific, technical and administrative Optics System” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
centre of ESO where technical develop- S. BAGNULO ET AL. The UVES Paranal Observatory Project . . . . . . . . . . . . . . . .10
ment programmes are carried out to pro- D. BONACCINI ET AL. Pushing Technologies: ESO Fibre Laser Development for
vide the Paranal and La Silla observato-
ries with the most advanced instruments. Laser Guide-Star Adaptive Optics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15
ESO employs about 320 international M. MAYOR ET AL. Setting New Standards with HARPS . . . . . . . . . . . . . . . . . . .20
staff members, Fellows and Associates in L. GERMANY News From La Silla . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25
Europe and Chile, and about 160 local
staff members in Chile.
REPORTS FROM OBSERVERS
The ESO MESSENGER is published four
times a year: normally in March, June,
A. RICHICHI & F. PARESCE Harvesting Scientific Results with the VLTI . . . . . . . .26
September and December. ESO also J. ALVES & N. HOMEIER Our Own Starburst . . . . . . . . . . . . . . . . . . . . . . . . . . . .35
publishes Conference Proceedings, Pre- R. TÜLLMANN ET AL. Extraplanar Star Formation in NGC 55 . . . . . . . . . . . . . . . .39
prints, Technical Notes and other material
connected to its activities. Press Releas-
ESO PRESS RELEASE Flares from the Direction of the Black Hole
es inform the media about particular in the Galactic Centre . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .43
events. For further information, contact ESO PR PHOTO 31/03 Roses in the Southern Sky . . . . . . . . . . . . . . . . . . . . . .44
the ESO Education and Public Relations
Department at the following address:
B. LEIBUNDGUT, U. GROTHKOPF & A. TREUMANN Metrics to Measure ESO’s
Scientific Success . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .46
EUROPEAN
SOUTHERN OBSERVATORY OTHER ASTRONOMICAL NEWS
Karl-Schwarzschild-Str. 2
D-85748 Garching bei München G. HAU & D. ALLOIN Report on the Topical Meeting “Resolved Stellar
Germany Populations” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .50
Tel. (089) 320 06-0
Telefax (089) 3202362
P. SHAVER ESO-ESA Meeting “Cosmology and Fundamental Physics” . . . . . .50
Email: [email protected] W. BRANDNER & M. KASPER ESO Workshop on “Science with Adaptive
URL: http://www.eso.org Optics” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .51
H.U. KÄUFL & A. MOORWOOD ESO Workshop on “High Resolution Infrared
The ESO Messenger: Spectroscopy in Astronomy” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .52
Editor: Peter Shaver A. RENZINI ESO-USM-MPE Workshop on “Multiwavelength Mapping of
Technical editor: Henri Boffin
http://www.eso.org/messenger/
Galaxy Formation and Evolution” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53
The VT-2004 Educational Programme - A Unique Opportunity . . . . . . . . . . . .55
Printed by B. MACKOWIAK & R. WEST Physics on Stage 3: Life and New Frontiers . . . . . .56
Universitätsdruckerei
WOLF & SOHN
Fellows at ESO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57
Heidemannstr. 166
D-80939 München
Germany ANNOUNCEMENTS
Future ESO Workshop on Planetary Nebulae beyond the Milky Way . . . . . . . .58
ISSN 0722-6691 ESO Vacancy : Infrared Astronomer, Head of the IR Instrumentation Dept. . . .59
Personnel Movements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59

60 The Messenger 114