messenger-no119

No. 119 – March 2005

2 The Messenger 119

SSCIENTIFICCIENTIFIC SSTRATRATEGYTEGY PPLANNINGLANNING AATT ESOESO

At its 100th meeting in June 2003, the ESOCouncil decided to install a Working Groupto discuss ESO’s scientific strategy until2020. The time appeared ripe to discuss the future, as the VLT was now largely com-pleted, ALMA had just been approved, and the ESO community had been significant-ly strengthened by the recent accession of the United Kingdom. Furthermore, the dis-cussions and concept studies for the nextlarge facilities (notably the Extremely LargeTelescopes) were underway world-wide. Evi-dently, it was important to develop a strategyfor ESO’s future now.

The Scientific Strategy Working Groupwas composed of members of Council: Ralf

INTRODUCTION

1. ESO’s mission was stated in the Con-vention as to “establish and operate anastronomical observatory in the southernhemisphere, equipped with powerful instru-ments, with the aim of furthering and orga-nising collaboration in astronomy”. In thecurrent world and in view of Europe’s andESO’s achievements in astronomy in the lastdecade, ESO’s mission could be stated moreambitiously: ESO should provide Europeanastronomers world-class facilities to pursuethe most fundamental astronomical ques-tions.

2. ESO cannot do this alone. A closepartnership between ESO and the astronomi-cal institutions in its member countries is cru-cial to the development and preservation of the scientific and technical excellence of European Astronomy. This implies that the success of European Astronomy reliesequally on a world-class ESO and on strong and active research institutions throughoutEurope.

3. The granting of access to ESO facilities,participation in ESO programmes, and evenmembership of ESO are based primarily onscientific excellence. ESO will continue to beopen to new members and collaborations, fol-lowing the principles of furthering excellenceand scientific cooperation.

THE ASTRONOMICAL FRAMEWORK

4. Over the past two decades, Astronomyhas entered its golden age. A few examplesof what has been achieved are: We have nowdirect evidence for the evolution of galaxies

Bender (Chair), Tim de Zeeuw, Claes Frans-son, Gerry Gilmore and Franco Pacini; BrunoMarano and members of the STC, the VLTI Implementation Committee and theEuropean ALMA Board: Jean-Loup Puget,Thomas Henning, and Simon Lilly. ESO wasrepresented by Bruno Leibundgut, Guy Mon-net and Peter Quinn. The Director Generaland the Head of Administration attended allmeetings as well.

The charge to the Working Group wasdefined by Council as follows: “Prepare andassess the options for ESO’s long term pro-gramme, taking a broad view of ESO’s rolein world astronomy [...]. In doing so, theGroup shall consider ESO’s long term scien-

tific goals and objectives. To this end, currentand future developments and the possibleimplications of further external collaborationand enlarged membership may also be con-sidered.”

The Working Group met three times andprepared a report accompanied by a set ofrecommendations. After minor revisions, theESO Council adopted the recommendationsas a formal “Council Resolution on ScientificStrategy” in its last meeting in December2004. Both documents are printed below.

I take this opportunity to thank all Work-ing Group members for very good and opendiscussions, constructive contributions andpleasant and efficient meetings.

INTRODUCTION BY RALF BENDER, CHAIR OF THE SCIENTIFIC STRATEGY WORKING GROUP

REPORT FROM THE WORKING GROUP ON SCIENTIFIC STRATEGY PLANNING

and stars throughout 90 % of the age of theuniverse. We have found supermassive blackholes in most galaxy centres and have probedtheir evolution to high redshifts. We haveseen the seeds of galaxies and their large-scale distribution in the cosmic microwavebackground. We have determined the cosmicparameters with an order of magnitude betteraccuracy. We have confirmed the existenceof dark matter which is 5 times more abun-dant than ordinary matter, and we have foundthat the universe is filled to 70 % with the so-called ‘dark energy’, a new state of ener-gy of hitherto unknown nature. And last, notleast, we have, for the first time, found plan-ets around other stars.

5. It is evident that progress in astronomyis driven by both unexpected discoveries (e.g. dark matter, dark energy) as much as by particular experiments designed to testspecific theories (like ongoing microwavebackground experiments). Astronomical dis-coveries are often made by pushing the lim-its of observation with the most powerful tele-scopes on the ground and the most advancedsatellites in space (e.g. high redshift galax-ies, black holes, gamma ray bursts etc), butsmaller workhorse telescopes and instru-ments used in a new mode of operation canalso produce very exciting discoveries (e.g.MACHOS, planets etc.).

6. Key scientific questions in astronomyand astrophysics over the next 20 years willinclude (i) the nature of dark matter, (ii) thenature of dark energy, (iii) the formation ofthe very first stars and galaxies and follow-ing their evolution from the highest redshifts

until today, (iv) extreme conditions of matterand energy (e.g. black holes), (v) the forma-tion of stars and planetary systems, and (vi)the characterization of extra-solar planets in-cluding the search for extraterrestrial life.

7. Addressing all of these questions re-quires a co-ordinated observational and ex-perimental approach, spanning all wave-lengths of the electromagnetic spectrum withfacilities on the ground and in space but alsoexploring new observing windows to the uni-verse, like underground neutrino detectors, or space interferometers for gravitationalwave detection. Ground-based astronomywith large aperture telescopes plays a pivotalrole in the overall concept because (a) mostsources we want to study emit a large frac-tion of their radiation between optical andradio wavelengths, (b) this wavelength rangeprovides crucial and detailed informationabout the physical nature of the sources, and(c) the sources we want to study are general-ly very faint. In addition, large telescopes likeKeck and VLT have not only made importantdiscoveries by themselves but also haveprovided crucial complementary informationto discoveries made with satellites (e.g. the Hubble Space Telescope, ISO, Chandra etc.) which otherwise could often not be inter-preted comprehensively and would remaininconclusive. And finally, beyond their large light collecting power, the additionalstrengths of ground-based telescopes aretheir high versatility and the possibility toexplore new technologies rapidly.

8. Astronomy has from its very beginningbeen a technology-enabled science and is

3© ESO – March 2005

now progressing more rapidly than everbefore, with its technology feedback benefit-ing industry. In ground-based astronomy,improving the performance of existing tele-scopes and the construction of the next gen-eration of telescopes and instruments willrequire investment in several critical tech-nologies. Important over the next 10+ yearswill be, e.g., the development of multi-con-jugate adaptive optics, laser guide stars, the mastering of increasingly complex tele-scope/instrument systems and the handlingand exploration of Petabytes of data.

ESO’S FACILITIES

9. ESO and its collaborating instituteshave a highly skilled and very motivated staffspecialized in the design, construction andoperation of large optical / IR telescopes andtheir instruments. Another major strength ofESO is the efficient management of largeprojects which is one reason why the VLT isthe best ground-based astronomical facilitytoday. It took decades to build this expertiseand this asset must be preserved if Europe isto stay competitive in the future. ESO has alsomanaged its facilities effectively, opening anew site for the VLT because it was scien-tifically advantageous to do so, and closingfacilities on La Silla when no longer scien-tifically cost effective.

10. La Silla is still one of the mostsuccessful observatories world-wide. Surveyand monitoring projects with dedicated in-struments (e.g. HARPS for planets) havebecome increasingly important and produceimpressing scientific results. La Silla has also been needed to provide targets for the VLT. However, with the installation ofVST/OmegaCAM and VISTA, preparatoryobservations and target finding for the VLTwill not have to rely on La Silla beyond 2006.

11. The VLT is the most powerful and ver-satile 8 m telescope system to date. It is nowfully operational. Further upgrades and thedevelopment of second generation instru-ments should ensure European leadership in most areas of optical/IR astronomy for at least 10 more years. Once a 30 m+ tele-scope and the James-Webb-Space-Telescopego into operation, the role of the VLT needs

to be reconsidered and more specializationmay be required.

12. The VLTI is acknowledged to be themost advanced interferometer in the world in almost all aspects (except nulling interfer-ometry with Keck). It will be the best sys-tem to enable faint science (e.g. structure ofActive Galactic Nuclei) and ground-basedastrometry, because it is the only system that can potentially combine four 8 m tele-scopes interferometrically. The VLTI is stillbeing constructed, with new instruments tobe added and four Auxiliary Telescopes to becompleted.

13. ALMA will open a new window to theuniverse and provide unprecedented accessto the gaseous medium and the star formationprocesses both in our Galaxy and in the mostdistant galaxies in the universe. ALMA willdiscover vast numbers of faint sources thatrequire complementary observations at otherwavelengths.

14. These facilities, and many othersaround the world, produce an enormousamount of archived data which is avail-able to the astronomical community. ESO isworking with European institutions in a glo-bal effort to establish an International Virtu-al Observatory. This project addresses criti-cal requirements for handling the steadilyincreasing data rates from ESO telescopesand for connecting them with data setsobtained by other facilities and at other wave-lengths. Data handling and processing is oneof the key new challenges in astronomy.

THE EXTREMELY LARGE TELESCOPE

15. The unique capabilities of an Extreme-ly Large Telescope (30 m and larger) are (i) its 16 to 160 times larger light collectingpower than a VLT UT and (ii) its potentially10 to 40 times higher spatial resolution thanthe Hubble Space Telescope. The combina-tion of these two features will enable imag-ing and especially spectroscopy of sources upto a factor 50 fainter than currently possi-ble. Considering the enormous progress the Hubble Space telescope brought with its fac-tor 10 improvement in spatial resolution, andthe 8 m class telescopes with their factor 5 inlight collecting power, the discovery powerand impact of an ELT can hardly be overes-timated. With an appropriate choice of the sitean ELT should also offer unique imagingcapabilities in the sub-mm range comple-menting ALMA.

16. The case for an ELT alone is compel-ing; in the context of other facilities it is over-whelming. An ELT will be an important com-plement to ALMA, the James Webb SpaceTelescope (JWST), future space missions likeDARWIN and XEUS, and to other space andground observatories. It is important to real-ize that, because of their intrinsically differ-ent capabilities, ELTs on the one hand andspace missions like DARWIN or JWST onthe other hand, will not compete but rathersupport each other by providing complemen-tary information about planets, stars and gal-axies. In combination, these facilities willproduce the next revolution in our under-standing of the universe and its constituents.

17. The scientific reach of an ELT, and itspotential for making new discoveries, are sogreat that there is a strong case for each of theworld’s regions having access to an ELT. Col-laborative efforts should be encouraged, butshould not be allowed to compromise Euro-pean access to an ELTor to its associated tech-nological benefits. It is therefore importantthat European astronomy builds on its currentstrength and aims for a leading role in thedevelopment and construction of an ELT.This is also vital in attracting the best youngscientists and keeping them in Europe. NorthAmerican institutions (CalTech, UC, AURA,Canada) are undertaking a detailed design

View of La Silla fromthe 3.6 m telescope.

The VLT with stations of the VLTI(foreground).

4 The Messenger 119

study of a 30 m telescope, the TMT. Anothergroup of institutions (led by Carnegie Ob-servatories) has started constructing a 21 mtelescope, the GMT. Their ambition is to havefirst light before 2016 for both projects.Europe must keep pace with this work.

18. ESO and European Institutions arejointly pursuing technology development andconcept studies towards an ELT (in partthrough the FP6 framework). ESO has devel-oped what appears to be the most innovativeconcept to date for an Extremely Large Tele-scope, namely the OWL1. The OWL conceptstudies have been carried out in close col-laboration with industry and indicate that afundamentally new approach to build largetelescopes should allow the construction of a 60 m telescope for a cost comparable to that of a conventional 30 m telescope. Thenew paradigm is based on the adoption of serialised industrial production and a fullycomputer-controlled optical system to reducecost without compromising performance.Below about 60 m, the OWL concept prob-ably loses its high cost effectiveness. Adetailed design study is essential to validatethe OWL approach, including instrumen-tation, and establish the optimal balancebetween science, technology, and cost.

19. The total cost for a 100 m ELT basedon the OWL concept is currently estimated tobe about 1200 M Euros. A 60 m ELT wouldcost about half this amount, or roughly thecost of VLT or ALMA.

20. Three illustrative scenarios have beendeveloped by ESO. They are not yet opti-mised for cash flow or resource usage, but aresufficient to illustrate the main points in theplanning. While still including some allo-cation for technological development incrucial areas, Scenario I corresponds to thefastest schedule technology could plausiblyallow. Scenario II-100 allows more extensivedesign and development periods before startof construction, and a relaxed integrationschedule. Scenario II-60 corresponds to a60 m instead of a 100 m telescope.

CONCLUSIONS

21. Over the last decade ESO has suc-ceeded in becoming fully competitive andindeed world leading. However, the risk offalling back is real, even in the near future,especially with respect to ELTs. To maintain its position Europe has to adopt plans which keep its scientific and technical ambitions at

1 In this document, the term ELT refers to all tele-scopes larger than 30 m, without reference to a specific design. Most technology developmentthat has been carried out up to now (e.g. withinEU-funded projects) is at the component leveland is indeed design independent. The term OWLis used only when referring to the specific tele-scope concept that has been developed by ESO.All the statements related to OWL in this docu-ment should be revisited after the conceptual de-sign review which is expected to take place at theend of 2005.

the highest level to remain attractive for thebest scientists and engineers. This means thatEurope, and specifically ESO, must partici-pate in the most important and challengingtechnical and scientific developments of thefuture and should set priorities accordingly.This is also important for European indus-tries.

22. In the different scenarios for ESO’sfuture until 2015, optimal support for the VLTand ALMA and the development of an ELTare unquestionable priorities. The operationof La Silla and the enhancement of VLTI canhave different priorities.

23. The success and excellence of Euro-pean astronomy requires ESO to maintain the VLT as a world-leading facility for at least10 more years. The VLT needs constant up-grading, including MCAO and an adaptivesecondary, and a vigorous 2nd generationinstrument program because ESO must: a. keep pace with the steadily improving

capabilities of other 8 m telescopes (Keck,Gemini, Subaru, LBT …),

b. continue to utilise technological ad-vances,

c. match the evolving European science re-quirements, and

d. maintain developments that are critical foran ELT.24. The unique capabilities of the VLTI

mean that the current generation of VLTIinstruments and PRIMA for two telescopesshould be completed with high priority. Thecase for the extension of PRIMAto four beamcombination using either 4 ATs or 4 UTs needsto be demonstrated with simulations of real-istic observing situations. If the demonstra-tion is convincing, four beam combinationshould be implemented.

25. The role of La Silla beyond 2006 hasbeen cogently presented. La Silla will still beuseful and competitive in many respects, butit will not be as essential for the success ofESO as the VLT, VLTI, ALMA and an ELT

which consequently must have higher prior-ity than the continued operation of La Silla.

26. The European components of ALMAare being constructed in collaboration withinstitutes within the European astronomicalcommunity. This collaboration is also criticalfor the development of adequate data analy-sis tools which will help to educate Europeanastronomers to make best use of ALMA andperform cutting-edge science projects. ESOshould build up sufficient competences in themm and sub-mm fields to coordinate andcomplement the expertise and support fromoutside institutions.

27. ESO must ensure that Europe pre-serves its current world-leading position into the ELT era, because ESO and Europeanastronomers cannot afford to be left be-hind in the most important developments in ground-based optical/IR astronomy. Thiscan be achieved through the construction of a 60 m OWL for a cost comparable to a US30 m telescope. Thanks to its new conceptbased on serialised production, an OWL-typetelescope can be realized at much lower costthan a ‘conventional’30 m telescope of Keck-design. However, the advantages of serialisedproduction only become effective beyondabout 60 m and are in fact being validated for a 100 m telescope.

28. One of the advantages of the ESOconceptual design is that the telescope is de-signed so that it can be ‘staged’ in diameter,becoming available for observations withonly a partially completed primary mirror. AnOWL-type telescope could have first-light asa 30 m telescope on a competitive timescaleand then grow to a 60 m over several addition-al years (the ALMA project already adoptsthe same philosophy), and similarly a 100 mtelescope could start operations as a 50 m or60 m. Similarly, Europe can stay competi-tive in timescale by adopting this ‘growing atelescope’ concept and so having access to a world class 30 m telescope at the same time

Scenario Telescope Phase B Phase C/D Start of science Full completiondiameter starts starts (partially filled)

I 100 m 2005 2007 2012 (50 m dia.) 2016

II -100 100 m 2006 2010 2017 (60 m dia.) 2021

II -60 60 m 2006 2010 2016 (40 m dia.) 2020

Artist’s image of theAtacama Large Milli-meter Array (ALMA).

5© ESO – March 2005Scientific Strategy Planning at ESO

as the North American colleagues. Strongcommunity involvement for system leveldevelopment and provision of instrumenta-tion will, as in the case of VLT and ALMA,be crucial for our success.

29. As with ALMA, a collaboration withNorth America and possibly others wouldenable an even more ambitious global ELTproject to be undertaken.

30. ESO should continue to develop newdata archiving, data access, and data mining

technologies in collaboration with the Euro-pean astronomical community and within theframework of the International Virtual Ob-servatory Alliance. The availability of thesetechnologies is an important factor for thefuture success of ESO and European Astro-nomy.

31. Because astronomy and astrophysicsexploit leading edge technology, ESO shouldremain at the forefront of future mainstreamand key technologies concerning telescopes,

– the VLT will continue to receive effec-tive operational support, regular upgrad-ing (especially to keep it at the forefrontin image quality through novel adaptiveoptics concepts) and efficient 2nd genera-tion instrumentation in order to maintainits world-leading position for at least tenmore years,

– the unique capabilities of the VLTI will beexploited,

– the construction of an Extremely LargeTelescope on a competitive time scale willbe addressed by radical strategic planning,especially with respect to the developmentof enabling technologies and the explo-ration of all options, including seekingadditional funds, for fast implementation,

– ESO and its community will continue theirsuccessful partnership and seek effectiveintercontinental collaborations in devel-oping the most important and challengingtechnologies and facilities of the future.

Artist’s impression ofESO’s OWL concept.

ESO COUNCIL RESOLUTION ON SCIENTIFIC STRATEGY

ESO Council, considering the report of itsWorking Group for Scientific Strategic Plan-ning, ESO/Cou-990, and its recommenda-tions in ESO/Cou-964 rev. 2, agrees that

– astronomy is in a golden age with newtechnologies and telescopes enabling animpressive series of fundamental dis-coveries in physics (e.g. dark matter, darkenergy, supermassive blackholes, extraso-lar planets),

– over the last decade, the continued invest-ment of ESO and its community into theimprovement of ground-based astronom-ical facilities has finally allowed Europeto reach international competitiveness andleadership in ground-based astronomicalresearch,

– the prime goal of ESO is to secure this sta-tus by developing powerful facilities inorder to enable important scientific dis-coveries in the future,

– only the continued investment in cuttingedge technologies, telescopes, instrumentsand information technology will enablesuch scientific leadership and discoveries,

– ESO will continue to be open to newmembers and collaborations, followingthe principle of furthering scientific excel-lence,

and accordingly adopts the following princi-ples for its scientific strategy:

– ESO’s highest priority strategic goal mustbe the European retention of astronomicalleadership and excellence into the era ofExtremely Large Telescopes by carefullybalancing its investment in its most impor-tant programmes and projects,

– the completion of ALMA is assured andconditions for an efficient exploitation ofits superb scientific capabilities will beestablished,

instruments, and data handling. To achievethis goal, ESO should continue its verysuccessful partnership with European astro-nomical institutions and industries as in thepast, also within the framework of EU fund-ed projects.

Tele

scop

esan

dIn

stru

men

tatio

n

6 The Messenger 119

TTHE DEVELOPMENT OF EVER-larger format infrared detec-tors with excellent uniformity,quantum efficiency and noiseperformance has made infrared

imaging a central tool in modern astronomi-cal research. As an example, Figure 1 showsthe steady increase in the amount of VLT-ISAAC near-infrared imaging time since itscommissioning, to its current level of around26 runs or 300 hours per observing period. Inaddition to ISAAC, NAOS-CONICA on theVLT and SOFI on the NTT also provide nearinfrared imaging. The reasons for the strongdemand are many and varied. Fundamental-ly, the infrared allows the study of astronom-ical phenomena in otherwise inaccessibleregions of time and space.

On a cosmological scale, galaxies withinthe redshift range 1.5 < z < 4 have their rest-frame visible wavelengths shifted to the nearIR. This then becomes the natural wave-length in which to study them, allowing adirect comparison with local galaxies. Indeedbroadband IR colours allow their distances(redshifts) to be estimated photometrically.

Searches in the infrared for extra high red-shift (z > 6) young galaxies are underway atthe VLT. Nearby galaxies benefit from IRimaging which reveals the older stellar popu-lation, less obscured by dust.

Closer to home, the star forming regionswithin our own galaxy are often hidden bydust. So in order to study important aspectsof young clusters such as the initial massfunction, infrared imaging is necessary topenetrate the dust, if a complete census ofobjects is to be compiled.

The infrared part of the spectrum also con-tains major emission lines. Perhaps the mostimportant of these are due to quadrupole tran-sitions of molecular hydrogen, the most com-mon form of hydrogen in dense clouds. Thisline, usually shock excited, can reveal spec-tacular large-scale outflows from young stars.

Another important advantage of the near-infrared is the better image quality that isachieved compared to visible wavelengths.Since the size of the seeing disc has an inverseone-fifth power law dependance on wave-length (depending somewhat on the atmos-pheric turbulent outer scale), images of point

sources at K-band can be 20% smaller thanin the visible and consequently sharper.

HAWK-I: AN OPTIMUM VLT IMAGER

Infrared detectors are expensive – around 10 times the cost of comparably sized CCDs.So achieving both adequate image samplingand an ambitious field of view tends to require a large number of detectors and has histor-ically been difficult. However, thanks to re-cent developments in IR detector technologywhich reduce the cost per pixel, the situa-tion has greatly improved. HAWK-I, the HighAcuity Wide-field K-band Imager, will be anear-optimum camera for the VLT. Table 1shows the key instrumental parameters. The 7.5 arcminute square field results in outercorners of the HAWK-I field which willencroach slightly into the vignetted area ofthe Nasmyth field, resulting in 0.4% lowerthroughput in the field corners in all bands,and an approximately 25% higher back-ground flux there in K-band. So this is practi-cally the largest IR field possible at Nasmythwhile keeping reasonably uniform sensitivi-ty in all bands. By then assembling a mosaic

WWIDEIDE FF IELDIELD IINFRAREDNFRARED IIMAGINGMAGING ONON THETHE VLVLTTWITHWITH HAHAWK-IWK-I

HAWK-I IS A NEW WIDE FIELD INFRARED CAMERA UNDER DEVELOPMENT AT ESO. WITH A 7.5 ARCMINUTE

SQUARE FIELD OF VIEW AND 0.1 ARCSECOND PIXELS, IT WILL BE AN OPTIMUM IMAGER FOR THE VLT, AND A

MAJOR ENHANCEMENT TO EXISTING AND FUTURE INFRARED CAPABILITIES AT ESO.

MMARKARK CCASALIASALI , J, JEANEAN-F-FRANÇOISRANÇOIS PPIRARDIRARD, M, MARKUSARKUS KKISSLERISSLER-P-PAATIGTIG , A, ALANLAN MMOOROORWOODWOOD,,

RROLLIOLLI BBEDINEDIN , P, PETERETER BBIEREICHELIEREICHEL , B, BERNARDERNARD DDELABREELABRE, R, REINHOLDEINHOLD DDORNORN,,

GGERERTT FF INGERINGER, D, DOMINGOOMINGO GGOJAKOJAK, N, NORBERORBERTT HHUBINUBIN , G, GOTTHARDOTTHARD HHUSTERUSTER, Y, YVESVES JJUNGUNG,,

FFRANZRANZ KKOCHOCH, M, MISKAISKA LLEE LLOUARNOUARN, J, JEANEAN-L-LOUISOUIS LL IZONIZON, L, LEANDEREANDER MMEHRGANEHRGAN,,

EESZTERSZTER PPOZNAOZNA, A, ARMINRMIN SSILBERILBER, B, BARBARAARBARA SSOKAROKAR, J, JÖRGÖRG SSTEGMEIERTEGMEIER

EUROPEAN SOUTHERN OBSERVATORY

Figure 1: Statistics ofnear-infrared ISAACimaging as a functionof observing period.The red line shows theallocated number ofruns, and the black linethe number of hours,per period.

Figure 2: The galaxyM104 (Sombrero) withsuperimposed fields of HAWK-I (7.5() andISAAC (2.5() in red, andFORS1 (6.8() in white.


of four 2k × 2k detectors to fill this field, apixel scale of 0.1 arcsec/pixel results, whichis sufficiently small to adequately sample the best seeing at Paranal, even with futureground layer adaptive optics correction. Theend result is an imager with the best possibleperformance, limited predominantly by thetelescope design and atmospheric seeing con-ditions. The enhanced field of view comparedto ISAAC is shown in Figure 2.

HAWK-I, VISTA AND KMOSESO users will also have access to imag-ing data from the 4 m VISTA IR camera(Emerson et al., 2004), which should be com-missioned in Chile on a similar timescale(2007) giving ESO astronomers enormousinfrared imaging power. With its 16 2k × 2kRaytheon detectors and 0.34 arcsec pixels,VISTA will cover 0.6 sq. degrees in a singleexposure, and will be a natural pathfinder forHAWK-I and other VLT instruments. Pecu-liar, interesting or clustered objects discov-ered with VISTAwill become targets for deepimaging and small mosaics at higher spa-tial resolution with HAWK-I/VLT. The twoinstruments will complement each other verywell.

HAWK-I is also expected to be a majorcontributor of targets for infrared multi-object integral field spectroscopy with thesecond-generation KMOS instrument, whichhas a comparable field but no imaging mode.

WHAT IS SPECIAL ABOUT THE

HAWK-I DESIGN?Although HAWK-I is a relatively simple im-ager, there are ambitious and novel aspects in the design which will enhance its perfor-mance. At the heart of the instrument are itsdetectors. HAWK-I will use four RockwellHawaii 2RG arrays to make its impressivefocal plane. These new-generation detectorsoperate from 0.8 to 2.5 microns with excel-lent uniformity and low dark current. Theyare also three-side buttable allowing a com-

pact focal plane with a cross-shaped gap of 2.7 mm or 15 arcsec. The detectors will beassembled in a package developed by GLSci-entific which allows all 32 channels per de-tector to be read out. A CAD drawing of theassembly is shown in Figure 3.

A unique aspect of HAWK-I will be itsvery high throughput. This is achieved witha powered window and all-reflective designto achieve an optics-only throughput of 90%.The layout is shown in Figure 4. The windowforms a pupil image at M3 which is the sys-tem cold stop. The high throughput alone will give HAWK-I a signal-to-noise improve-ment of 10–20% over other typical imagerssuch as ISAAC. The window is very large, at 404 mm of clear aperture, and is made ofinfrared-grade fused silica. Windows this sizecan suffer from potential frosting as the cen-tre cools by radiating into the cold cryostat,

Figure 3: Drawingshowing the HAWK-Ifocal plane, consist-ing of four coplanarRockwell Hawaii-2RGdetectors.

Figure 4: This cutawaydrawing shows theoptical components of HAWK-I and parts of the surround-ing cryostat.

Figure 5: This draw-ing shows HAWK-Iattached to the Nas-myth adapter (blue).The cable de-rotator(brown) is also shown.

while conductive coupling to the warm edgeis poor. Special care is being taken with baf-fle design and an emergency warm-air sup-ply to ensure that condensation does not oc-cur during operation.

HAWK-I has two six-position filter wheelsfor a total of 10 useable filter and two openpositions. Darks will be obtained by select-ing two different narrowband filters in eachwheel. The filter selection has been one of thetasks of the Instrument Science Team chairedby Adriano Fontana (Monte Porzio). Thefinal selection is shown in Table1. Apart fromthe usual broad and narrow-band filters, notethe methane-band filter for detection of coolbrown dwarfs, and three cosmological nar-row-band filters for detection of redshiftedLyman and hydrogen alpha emission lines.HAWK-I will attach to the Nasmyth adapteras shown in Figure 5.

Table 1: The key HAWK-I parameters.

Detectors 4 × 2k × 2k

Pixel scale 0.106)

Field of view 7.5( × 7.5(

Optical image quality < 0.2 arcsec at 80 %(excluding seeing) encircled energy

Optics-only 90%throughput

End-to-end system 50%throughput

Number of filter 10positions

Wideband filters Y, J, H, Ksordered

Rest-wavelength 1.58 (CH4)narrow-band filters 2.167 (Brγ)(microns) 2.122 (H2)

Cosmological 1.061narrow-band filters 1.187(microns) 2.090

8 The Messenger 119

PROJECT PROGRESS

The HAWK-I project completed its Final De-sign Review on November 17th 2004, and isnow entering the main manufacturing phase,although procurement of some long-lead timeitems, such as optics and detectors, has beenunderway for some time. The beginning ofassembly and integration should be in Sep-tember this year, leading to a Preliminary Ac-ceptance in Europe in mid-2006, and Provi-sional Acceptance Chile at the end of 2006.

HAWK-I AS A FIRST-LIGHT INSTRUMENT FOR

A VLT ADAPTIVE SECONDARY MIRROR

ESO is currently studying the possibility of equipping HAWK-I with a Ground Lay-er Adaptive Optics (GLAO) system calledGRAAL (Arsenault et al., 2004). Of coursean AO correction over the 7.5 arcmin field ofview will not deliver diffraction limited

images. But as a minimum requirement, theAdaptive Optics must reduce the 50% encir-cled energy diameter by 15% in Y and 30%in Ks band, when the natural seeing is 1 arc-sec. The ultimate goal of the AO system is tocorrect the atmospheric turbulence such thatthe instrument resolution becomes the limit-ing factor. That is, the Adaptive Optics sys-tem will provide the equivalent image quali-ty to 0.2 arcsec seeing. This would impactvirtually all observing programmes with bet-ter sensitivity and spatial resolution.

The feasibility of a deformable secondarymirror and laser system for the VLT is cur-rently being investigated. The results of thisstudy, as well as the operational impact ofsuch a facility, will be reviewed in the thirdquarter of 2005, with a decision to proceedwith the development, including laser tomog-raphy, to be taken possibly at the end of 2005.

Although a GLAO capability would comewell after HAWK-I commissioning, the re-quirements for AO have been incorporatedinto the HAWK-I design already. These in-clude allowing sufficient weight budget andspace between the cryostat window and in-strument rotator for an AO module. Tip-tiltcorrection must be done with natural guidestars, and options exist to use either the on-chip guide star mode of the Hawaii-2RG de-tectors, or to have a separate NGS pickoffouside the instrument. A Conceptual DesignReview for the HAWK-I AO was held the dayafter the HAWK-I instrument PDR in Decem-ber 2004.

REFERENCESArsenault, R., Hubin, N., Le Louarn, M., Monnet,

G., Sarazin, M. 2004, The Messenger, 115, 11Emerson, J. P. et al. 2004, The Messenger, 117, 27

ESO’ESO’SS TTWOWO OOBSERBSERVVAATORIESTORIES MMERGEERGE

On February 1, 2005, ESO merged its two ob-servatories, La Silla and Paranal, into one.This move will help ESO to better manage its many and diverse projects by deployingavailable resources more efficiently whereand when they are needed. The merged obser-vatory will be known as the La Silla ParanalObservatory.

Catherine Cesarsky, ESO's Director Gen-eral, commented on the new development:“The merging, which was planned during thepast year with the deep involvement of all the staff, has created unified maintenance andengineering (including software, mechanics,electronics and optics) departments acrossthe two sites, further increasing the alreadyvery high efficiency of our telescopes. It is mygreat pleasure to commend the excellent workof Jorge Melnick, former director of the LaSilla Observatory, and of Roberto Gilmozzi,the director of Paranal.”

La Silla, north of the town of La Serena,has been the bastion of the organization’sfacilities since 1964. It is the site of two ofthe most productive 4 m class telescopes inthe world, the New Technology Telescope(NTT) – the first major telescope equippedwith active optics – and the 3.6 m, which hosts HARPS, a unique instrument capable

of measuring stellar radial velocities with anunsurpassed accuracy better than 1 m/s, mak-ing it a very powerful tool for the discoveryof extra-solar planets. In addition, astrono-mers have also access to the 2.2 m ESO/MPGtelescope with its Wide Field Imager camera.Moreover, the infrastructure of La Silla is still used by many of the ESO member statesfor targeted projects such as the Swiss 1.2 mEuler telescope and the robotic telescope spe-cialized in the follow-up of gamma-ray burstsdetected by satellites, the Italian REM (RapidEye Mount). La Silla is also in charge of the APEX (Atacama Pathfinder Experiment)12 m sub-millimetre telescope which willsoon start routine observations at Chajnantor,the site of the future Atacama Large Millime-ter Array (ALMA). The APEX project is acollaboration between the Max Planck Soci-ety in Germany, the Onsala Space Observa-tory in Sweden and ESO.

Paranal is the home of the Very Large Telescope (VLT) and the VLT Interferometer(VLTI). Antu, the first 8.2 m Unit Telescopeof the VLT, saw First Light in May 1998,starting what has become a revolution in Eu-ropean astronomy. Since then, the three otherUnit Telescopes – Kueyen, Melipal and Ye-pun – have been successfully put into opera-

tion with an impressive suite of the mostadvanced astronomical instruments. The in-terferometric mode of the VLT (VLTI) is alsooperational and fully integrated in the VLTdata flow system. In the VLTI mode, onestate-of-the-art instrument is already avail-able and another will follow soon. In additionto the state-of-the-art Very Large Telescopeand the four Auxiliary Telescopes of 1.8 mdiameter which can move to relocate in up to30 different locations feeding the interferom-eter, Paranal will also be home to the 2.6 mVLT Survey telescope (VST) and the 4.2 mVISTA IR survey telescope.

Both Paranal and La Silla have a provenrecord of their ability to address the currentissues in observational astronomy. In 2004alone, each observatory provided data for thepublication of about 350 peer-reviewed jour-nal articles, more than any other ground-based observatory. With the present mergingof these top-ranking astronomical observato-ries, fostering synergies and harmonizing themany diverse activities, ESO and the entirecommunity of European astronomers willprofit even more from these highly efficientresearch facilities.

(based on ESO Press Release 03/05)


NNEWEW OBSEROBSERVINGVING MODESMODES OFOF NACONACOAFTER MORE THAN TWO YEARS OF REGULAR OPERATION, A NUMBER OF UPGRADES HAVE RECENTLY BEEN

INSTALLED FOR NACO AND OFFERED TO THE COMMUNITY. THE ARTICLE DESCRIBES THE NEW OBSERVING

MODES AND PROVIDES EXAMPLES OF ASTRONOMICAL APPLICATIONS IN HIGH-CONTRAST IMAGING, SPEC-TROSCOPY AND POLARIMETRY.

MARKUS KASPERMARKUS KASPER11, NANCY AGEORGES, NANCY AGEORGES22, ANTHONY BOCCALETTI, ANTHONY BOCCALETTI33,,

WOLFGANG BRANDNERWOLFGANG BRANDNER44, LAIRD M. CLOSE, LAIRD M. CLOSE66, RIC DA, RIC DAVIESVIES55, GER, GERT FINGERT FINGER11,,

REINHARD GENZELREINHARD GENZEL55, MARKUS HAR, MARKUS HARTUNGTUNG22, ANDREAS KAUFER, ANDREAS KAUFER22,,

STEPHAN KELLNERSTEPHAN KELLNER44, NORBER, NORBERT HUBINT HUBIN11, RAINER LENZEN, RAINER LENZEN44, CHRIS LIDMAN, CHRIS LIDMAN22,,

GUY MONNETGUY MONNET11, ALAN MOOR, ALAN MOORWOODWOOD11, THOMAS OTT, THOMAS OTT55, PIERRE RIAUD, PIERRE RIAUD33,,

HERMANN-JOSEF RÖSERHERMANN-JOSEF RÖSER44, DANIEL ROUAN, DANIEL ROUAN33, JASON SPYROMILIO, JASON SPYROMILIO22

1EUROPEAN SOUTHERN OBSERVATORY, GARCHING, GERMANY

2EUROPEAN SOUTHERN OBSERVATORY, SANTIAGO, CHILE

3LESIA, OBSERVATOIRE DE PARIS, MEUDON, FRANCE

4MAX-PLANCK-INSTITUT FÜR ASTRONOMIE, HEIDELBERG, GERMANY

5MAX-PLANCK-INSTITUT FÜR EXTRATERRESTRISCHE PHYSIK, GARCHING, GERMANY

6STEWARD OBSERVATORY, TUCSON, USA

NNAOS-CONICA (HEREAFTER

NACO) is a near-infraredimager and spectrograph fedby an Adaptive Optics (AO)system to correct for optical

aberrations introduced by atmospheric turbu-lence. NACO saw first light on November 25,2001, at VLT UT4 (Brandner et al. 2002,Lagrange et al. 2003, Lenzen et al. 2003) andhas been offered to the astronomical com-munity since October 2002 (period 70). Sincethen, the science output of NACO amountsto more than 30 refereed articles with sever-al highlights such as the confirmation of theblack hole in the Galactic Centre (Schödel etal. 2002) and discovery of its flares (Genzelet al. 2003), as well as the dynamical calibra-tion of the mass-luminosity relation at verylow stellar masses and young ages (Close etal. 2005, see Figure 1).

The AO system NAOS was built by a French consortium comprised of OfficeNational d’Etudes et Recherches Aérospa-tiales (ONERA), Observatoire de Paris andLaboratoire d’Astrophysique de l’Observa-toire de Grenoble (LAOG). It compensatesfor the effects of atmospheric turbulence(seeing) and provides diffraction-limitedresolution for observing wavelengths from 1 to 5 µm, resulting in a gain in spatial reso-lution by a factor of 5 to 15 (diffraction limitof an 8-m-class telescope in K-band corre-sponds to 60 mas). Active optical elements of

NAOS include a tip-tilt plane mirror and adeformable mirror (DM) with 185 actua-tors. NAOS is equipped with two Shack-Hartmann type wavefront sensors (WFS) forwavefront sensing at optical (450 to 950 nm)and near-infrared (0.8 to 2.5 µm) wave-lengths.

The near-infrared imager and spectro-graph CONICA was built by the Ger-man Max-Planck-Institutes für Astronomie(MPIA) and für Extraterrestrische Physik(MPE) and ESO. In its original configuration,CONICA already provided six cameras for imaging and long-slit spectroscopy in the near-infrared between 1 µm and 5 µm atvarious spatial and spectral resolutions on a1 × 1 kilopixel Aladdin detector, about 40 dif-ferent filters for broad- and narrow-band im-aging, Wollaston prisms and wiregrids forpolarimetry, various grisms and a cryogenicFabry-Perot interferometer for spectroscopy,as well as a Lyot-type coronagraph with dif-ferent mask diameters.

The great flexibility of the instrument con-cept triggered many ideas on how the capa-bilities of NACO could be extended and opti-mized for certain specialized astronomicalapplications. For example, NACO was lack-ing a differential imaging mode and an ade-quate coronagraphic mode for high-contrastobservations close to bright stars with theultimate goal of detecting extrasolar planets.Both concepts are new developments for

this kind of observations and have proven intheory to enhance the achievable contrast by orders of magnitude (Marois et al. 2000,Rouan et al. 2000). Additionally, new spec-troscopic and polarimetric modes have beenproposed with the main goal of increasingNACO’s efficiency and minimize time over-heads. Table 1 lists and briefly describes the recently offered modes and upgrades, ofwhich the major ones will be discussed in thefollowing sections of this article.

Figure 1: An enhancedfalse color infraredimage of AB Dor A andC. The faint compan-ion “AB Dor C” – seenas the pink dot at 8 o’clock – is 120 timesfainter than its primarystar. This is the faintestcompanion ever direct-ly imaged within 0.156)of its primary. Never-

theless, the new NACOSDI camera was able to distinguish it as aslightly “redder speck-le” surrounded by the “bluer” speckles from AB Dor A. It takes 11.75 years forthe 93 Jupiter masscompanion to complete this orbit shown as aorange ellipse.

10 The Messenger 119

NEW ALADDIN III DETECTOR

The most recent upgrade of NACO per-formed by the ESO IR department in collab-oration with MPIA was the replacement ofthe CONICA Aladdin II detector. At largerreverse bias voltages (high dynamic range)the cosmetic properties of the old Aladdin IIarray were degraded substantially by a largenumber of warm pixels. For this reason, it wasoperated close to full well applying a lowreverse bias voltage (for NACO users definedas high sensitivity mode) resulting in an ac-ceptable cosmetic appearance. However, thecharge storage capacity of the detector in thismode was small, and – as a property commonto all infrared detectors operating in capaci-tive discharge mode – the response was quitenonlinear close to the full well capacity. Thenew Aladdin III array promised substantialimprovements of the cosmetic properties.

After installation, the new detector be-haved as expected and showed greatly im-proved cosmetic properties. In addition, italso showed lower read-noise and flux zeropoints which are lower by 0.1–0.4 mag-nitudes, suggesting a higher quantum effi-ciency of the new array as summarized inTable 2.

SIMULTANEOUS DIFFERENTIAL IMAGER (SDI)Radial velocity searches provide evidencethat giant extrasolar planets are common. BySeptember 2004, 136 extrasolar planets hadbeen discovered and about 6% of the ob-served stars in the solar neighborhood hostextrasolar planets (see www.exoplanets.org).However, the radial velocity method is mostsensitive to planets in close orbit up to a fewAU and cannot measure parameters otherthan the planet’s orbit and mass with an uncer-tainty due to the unknown inclination of theorbit. Direct detection of extrasolar planets isrequired if we are to learn about the objectsthemselves. With direct detection we can ana-lyze photons directly from the companions

Table 1:NACO upgrades

Upgrade

SDI (simultaneousdifferential imager)

4QPM (Four quadrantphase mask)

Low-res prism

Order sorting filters SLand SHK

Fabry-Perot inter-ferometer

Superachromatic retarder plate

New Aladdin III detector

Offered since

P74

P74

P74

P74

P74

P75

05/2004

Date of installation

08/2003

01/2004

04/2004

04/2004

–

04/2004

05/2004

Short description

Uses a quad filter to take images simultaneously at 3 wavelengths surrounding the H-band methanefeature at 1.62 µm. Performing a difference ofimages in these filters reduces speckle noise andhelps to reveal faint methane companions next to bright stars.

Subdivides the focal plane in quadrants and delaysthe light in two of them by half a wavelength at2.15 µm. This coronagraphic technique helps tosuppress the light of a star in order to reveal faintstructures surrounding it.

Allows for simultaneous spectroscopy from J- to M-band at R = 50 to 400.

Allows for L-band and H+K-band spectroscopy at various spectral resolutions.

Allows for narrow band (2 nm) observations tun-able between 2 and 2.5 µm (for more informationsee Hartung et al. 2004).

Facilitates polarimetry by providing the possibility to rotate the position angle of polarization. Linearstructure can thus be observed over the entire FoVand observation overheads are reduced.

Replaces old Aladdin II detector and has bettercosmetics, linear range and read-noise.

allowing us to measure Lbol, Teff, SpectralType, and other critical physical characteris-tics of these poorly understood objects.

Young (100 Myr old) and massive (coupleof Jupiter masses to 12 Jupiter masses) extra-solar planets are 100000 times more self-lu-minous than old (5 Gyr) extra-solar planets,whereas their host stars are only slightlybrighter when this young. The luminositycontrast between them can thus be in therange 10–4 to 10–6 with the precise value de-pending on age and mass of the planet. Cur-rently the majority of such young stars thatare nearby (< 50 pc) are located in the south-ern star forming regions and associations(DEC < –20). To detect a faint planet near a bright star requires the high Strehl ratiosdelivered by NACO. However, NACO (likeall AO systems) suffers from a limiting“speckle-noise” floor which prevents the de-tection of planets within 1) of the primary star.Hence NACO required some method to sup-press this limiting “speckle noise” floor, soplanets could be imaged within 1) of their pri-mary star.

The SDI concept to reduce speckle noiseby L. M. Close from Steward Observatoryand R. Lenzen from MPIA (Lenzen et al.2004) is based on a method presented byMarois et al. (2000). This method exploits

the fact that all extra-solar giant planetscooler than about 1300 K have strong CH4

(methane) absorption past 1.62 µm in the H-band NIR atmospheric window, making theplanet virtually disappear at this wavelength.The difference of two images taken simulta-neously on either side of the CH4 absorptiontherefore efficiently removes star and speck-le noise while the planet remains. A criticalpoint of the design is the differential opticalaberration after the two wavelengths havebeen optically separated. It has to be kept verysmall (< 10 nm rms) in order to match thespeckle pattern at the two wavelengths andefficiently remove it by the image subtraction.

Figure 2 shows the optical concept of SDI.A double Wollaston prism, made of two iden-tical Calcite Wollaston prisms rotated againsteach other by 45 deg, splits the beam into four beams of equal brightness (for unpolar-ized objects). The f/40 camera images thefour beams onto the detector with a minimumseparation of 320 pixels or 5.5) at the pix-elscale of 17.2 mas per pixel. Currently, thefield of view is limited by a field mask andthe usable detector area to about 3) × 3.7).Just in front of the detector, each of the beamspasses through one of a set of narrow bandfilters with central wavelengths of 1.575,1.600 and 1.625 µm and a FWHM of 25 nm.

Figure 2: SDI opticalconcept. A doubleWollaston prism splitsthe beam into fourwhich are imagedthrough different filterslocated on the eitherside of the H-bandmethane feature at 1.62 µm.

11© ESO – March 2005Kasper M. et al., New observing modes of NACO

Pixels with variable dark current [%] Readout noise [ADU] Zero Points [mag]High sensitivity High dynamics High well depth Uncorr Fowler J H Ks

Aladdin II 0.72 7.1 56 5.6 2.1 24.01 23.87 22.99

Aladdin III 0.023 0.156 2.43 4.42 1.29 24.47 24.22 23.31

Table 2: Main propertiesof the old (Aladdin II)and the new (Aladdin III)Conica arrays.

Figure 4: Final SDIimage (double filtered)derived from real datawhich was used toderive the detectionlimits shown in Fig-ure 3. The three artifi-cial companions at0.5), 1) and 1.5) with∆mag = 9.5 (5σ detec-tion at 0.5)) are clearlyvisible.

The f/40 imaging optics was built such that itcould just replace the existing camera L100which had never been offered.

The contrast ratio achievable by SDI ismore than two magnitudes better than thatachieved by standard imaging and PSF sub-traction using a subsequently observed pointsource. Figure 3 shows the achievable con-trast for 5 sigma detection as a function ofangular separation for a 32 minute exposure.The highest contrast (“double filtered”) isachieved by subtracting the images taken attwo different rotator angles in addition to thedual image subtraction in order to reduce theremaining instrumental speckles and by fil-tering out the low spatial frequencies of theimages. SDI removes most of the specklenoise such that the contrast is mostly limitedby stellar photon noise at intermediate angu-lar separations and by sky background anddetector read-out noise at larger angular sepa-rations. In both cases, the achievable con-trast can be further improved by increasing the integration time. Figure 4 shows a re-duced double filtered SDI image. Compan-ions 9.5 magnitudes fainter than the star, areeasily detected outwards of 0.5).

Although mainly conceived for exoplanetimaging, SDI is also very useful for obser-vations of objects with thick atmospheres in the solar system like Titan. Peering at the same time through a narrow, unob-scured near-infrared spectral window in the dense methane atmosphere and an adjacent non-transparent waveband, Figure 5 showsTitan’s surface regions with very differentreflectivity in unprecedented detail whencompared to other ground-based observa-tions.

FOUR-QUADRANT PHASE MASK (4QPM)As for the SDI, the main scientific motivationfor the 4QPM coronagraph is to increase thecontrast of faint objects around bright stars.Besides the search for faint point sources asdescribed in the previous section, the 4QPMcan be used to look for hot dust around AGN(see e.g. Gratadour et al. 2005), quasar hostgalaxies or circumstellar emission producedby disks very close (0.1) to 0.5)) to the cen-tral point source.

The four quadrant phase-mask corona-graph was proposed by Rouan et al. (2000).The focal plane is split into four equal areas,two of which are phase-shifted by π. As aconsequence, a destructive interference oc-curs in the relayed pupil, and the on-axis star-light transferred outside the geometric pupilis blocked by a so-called Lyot stop. Theadvantage of the 4QPM over the Lyot maskis twofold: (1) no large opaque area at the cen-tre and an inner working radius of about

Figure 3: SDI detectivityas a function of angu-lar separation from the star using different data reductionschemes. ‘SDI doublefiltered’ is a techniquewhere the instrument isro-tated half waythrough the observationto calibrate instrumen-tal speckles with sub-sequent filtering of lowspatial frequency struc-tures. ‘l160’ denotesconventional imag-ing without speckleremoval.

Figure 5: Titan imagedwith SDI. The pictureshows Titan imagedthrough the 4 channelsof the SDI camera.Obviously, Titan ap-pears very faint and featureless whenimaged inside themethane band in thebarely visible lower two images. The cen-tral color image is the difference betweenthe out- and in-bandimages and has been added to the pic-ture afterwards.

3.4)


1 Airy disc, and (2) a larger achievable con-trast if good optical quality is met.

The actual concept of the 4QPM in NACO as proposed by LESIA, Observatoirede Meudon, consists of a SiO2 substrate witha 2.5 µm thick SiO2 layer deposited on twoof the quadrants. This device is placed in the CONICA mask wheel and has a workingwavelength of 2.15 µm where it achieves theπ phase-shift and maximum light rejection.The theoretically achievable PSF attenua-tion deteriorates with the square of the wave-length deviation from optimum. In practice,residual wavefront errors dominate overchromatic effects, and PSF core attenuationof about a factor 10 can be achieved all overthe K-band while it drops to a modest factorof 4 in H-band (Boccaletti et al. 2004). Fig-ure 6 displays the radial point source sen-sitivity achieved with NACO in a 10 minutesexposure observing a bright star. The actual-ly achievable 4QPM performance and thecontrast improvement by subtracting a subse-quently observed reference PSF star strong-ly depend on the quality and stability of theAO correction.

Figure 7 (top) shows a beautiful exampleof an astronomical application of the 4QPMpublished by Gratadour et al.(2005). The ob-servations show a complex environmentcloser to the nucleus than previously imagedat this wavelength. The identified structuresare similar to what has been observed previ-ously at longer wavelengths (3.8 and 4.8 µm),similar resolution, but without coronograph-ic mask. Up to now they were totally hiddenby the dominating emission of the nucleus atKs. Shape and photometry are in very goodagreement with the previous interpretation ofelongated knots, shaped by the passage of ajet, and composed of very small dust grains,transiently heated by the central engine of theAGN. On the bottom of the figure, the imageof a triple system HIP 1306 demonstrates theenhanced contrast for close companion detec-tion achievable with the 4QPM.

LOW-RESOLUTION PRISM

There are a number of research projects in which simultaneous, moderate resolutionspectro-photometry would be useful and, insome cases, essential. Perhaps the most press-ing and important case presently is the explo-ration of the infrared flares from the Galac-tic Centre black hole discovered with NACO.The flares promise to be a key tool for study-ing the physical processes in the strong grav-ity regime just outside the event horizon. Fora better understanding of the emission mech-anisms, it is necessary to obtain simultaneousspectral energy distributions (SED) acrossthe H-through L/M-bands. Since the flareslast typically only one hour and have timesubstructure of 10–20 minutes, it is not pos-sible to obtain the data sequentially. Subtletime variability of the SED, as expected if gasfalls in through the innermost accretion zone

Figure 7: Top: Struc-tures around NGC 1068revealed at both largeand close separations.The image to the right isPSF reference star sub-tracted. The contrast ofknown structures isimproved with respectto previous non-coro-nagraphic observations(Gratadour et al. 2005).

Bottom: Triple systemimaged with the 4QPM.The two companionsare at separations of0.128) and 1.075) withbrightness ratios to the main surpressedprimary of ∆m = 1.6and 3.5 mag (Boccalettiet al. 2004).

Figure 6: Radial detec-tivity of the 4QPM. The achievable 3σ de-tectivity is 10-4 (10 mag)at 0.5) and 10-5

(12.5 mag) at 1). Stellarresiduals, i.e., thequality of the PSF refer-ence subtraction, domi-nate at short angulardistance, while sky anddetector noise domi-nate at larger angularseparations for this starmagnitude and expo-sure time (courtesyAnthony Boccaletti).

and experiences increasing gravitational red-shift, also calls for simultaneous information.Other fields where the proposed new modewould be extremely useful are determinationof stellar types in dense star clusters, dust fea-tures in AGN environments and spectral char-acterization of brown dwarfs with deep andcharacteristic broad atmospheric IR features.

The concept of the low spectral resolu-tion mode of NACO has been developed by R. Lenzen from MPIA. SrTiO3/CsBr turnedout to be the best combination of a high dis-persion and a low dispersion material, pro-viding high transmission and a rather con-stant dispersion for the whole wavelengthregion. Figure 8 shows the spectral resolutionachievable with different material combina-tions over the spectral range of the CONICAdetector.

Ks+4QPM, fov = 11.7)

Large scale arclectsstructures (d = 160 pc) south tail

Ks+4QPM, fov = 3.5)

micro spiral arms (d = 15 pc)

The low resolution prism was used by theGalactic Centre Group of the Max-Planck-Institut für Extraterrestrische Physik in July2004 with the aim of measuring the spectralslope of the flaring source at SgrA* in theGalactic Centre. Since no flare was seen dur-ing this run, spectra of several stars in thecentral cluster were obtained instead, as a fea-sibility test and to better characterise the per-


formance of the prism. The adaptive opticsloop was closed on the infrared-bright IRS 7,using the infrared wavefront sensor. Due tothe high extinction towards the Galactic Cen-tre (AV = 30 mag), there is little to be seenshortward of H-band; and because the JHKdichroic was used for the infrared wave-front sensor (in order to be able to detect the LM-bands) a 2.5 µm short cutoff filter wasinserted. The resulting normalized LM-bandspectra of 3 stars are shown in Figure 9 (IRS7 with K ~ 6.7, IRS 16 NW with K ~ 9.5, andIRS 33 N with K ~ 10.5). Wavelength cali-bration was rather tricky and eventuallyachieved via a combination of narrow-bandfilter images and matching atmospheric ab-sorption features. No correction for extinc-tion has been applied to these spectra. Seeinglimited L-band spectra of several other Ga-lactic Centre stars have previously been pub-lished by Moultaka et al. (2004), and sim-ilar features can be clearly identified: H2Oice absorption at 3.0 µm; CH2 and CH3 ab-sorption at 3.4 µm; Pf γ and Brα emission at 3.74 µm and 4.05 µm respectively. The M-band data are the first such spectra of these Galactic Centre stars. The extraction ofa spectrum for at least one of the S-sources(not shown) indicates that it should certainlybe possible to obtain a 3–5.5 µm spectrum of a flare from SgrA* if one should be luckyenough to be using the prism at the right time. Actually, a K-band spectrum of a Galac-tic centre flare was obtained in July 2004 withthe adaptive optics based 3D IR spectrometerSINFONI during its first commissioning run.

SUPERACHROMATIC RETARDER PLATE

For the detailed study of extragalactic radiojets NACO is a unique instrument. Especial-ly in its polarization mode, information canbe collected which can be combined withradio data at the same resolution in order togain insight into the physics of these objects.Polarization studies in dusty star formingregions, circumstellar discs and envelopes,and scattering regions in AGN are other cen-tral areas where NACO is unparalleled.

Before the upgrade, polarimetry was pos-sible using two Wollaston prisms with posi-tion angles 0° and 45° with two main limita-tions: (1) Linear structures (like a jet) couldnot be covered at once over their entire length,if these objects were longer than the angu-lar separation of the Wollastons (3.3)). Thenmultiple images were necessary to cover theobject completely, as by rotation to e.g. 45°,part of the object will fall out of the stripemask. (2) With each rotation of the instrumentbetween 0° and 45° a new alignment of thetelescope pointing was necessary, leading toa substantial overhead in polarization obser-vations.

With the new turnable retarder plate at theinstrument entrance, just one Wollastonprism is used, and rotation of the plane ofpolarization is done by turning the plate.

ment of the system will be the installation of the Laser Guide Star Facility planned for spring/summer 2005. Additional compo-nents that NACO requires to operate with anLGS such as an additional STRAP tip tilt sen-sor as well as the necessary software upgradeshave already been installed and tested. Thefirst commissioning of NACO with the LGSis foreseen for fall/winter 2005. Then, a muchlarger fraction of astronomical objects will be accessible to the superb observing modesoffered by NACO.

REFERENCESBoccaletti, A. et al. 2004, PASP, 116, 1061Brandner, W. et al 2002, The Messenger, 107, 1Close, L. M. et al. 2005, Nature, 433, 286Genzel, R. et al. 2003, Nature, 425, 934Gratadour, D. et al. 2005, A&A, 429, 433Hartung, M. et al. 2004, Proc. SPIE, 5492, 1531Hartung, M. et al. 2004, A&A, 421, L17Lagrange, A. M. et al. 2003, Proc. SPIE, 4841, 860Lenzen, R. et al. 2003, Proc. SPIE, 4841, 944Lenzen, R. et al. 2004, Proc. SPIE, 5492, 970Marois, C. et al. 2000, PASP, 112, 91Moultaka, J. et al. 2004, A&A, 425, 529Rouan, D. et al. 2000, PASP, 112, 1479Schödel, R. et al. 2002, Nature, 419, 694

Figure 8: Spectral re-solving power over theCONICA NIR rangeprovided by prisms ofdifferent composition.

Kasper M. et al., New observing modes of NACO

The retarder plate was put in place of thenever offered TADC-device and is locatedabout 12 cm pre-focal within the converging f/15 beam. The resulting focus shift due to theachromatic plate thickness of about 8 mm(3.43 mm MgF2 thickness, 4.35 mm Quartzplate thickness) can be compensated by focusoffsetting NAOS.

Figure 10 shows an example of polarim-etry with NACO using one Wollaston prismand the retarder plate. The scientific goal ofthe observations was a polarization map ofthe jet of the quasar 3C273. The jet extendsout to about 20) from the quasar core locat-ed in the upper left. For this image, 94 expo-sures of 100 seconds each (NDIT = 1) havebeen added up. Only the common area of thefield is shown. The jet does seem to exhibit apolarization signal (most conspicuous at in-ner and outer knots). Because the quasar is sobright, the image also reveals reflections dueto the half wave plate and electronic ghostswithin the detector.

OUTLOOK

Although no further additional observingmodes are planned so far, the work on NACOhas not yet finished. The next major improve-

Figure 10 (below):Polarization data of the quasar 3C273 jet.NACO has been ro-tated to align the jetwith the Wollastonstripe mask. The retar-der plate can be usedto turn the polarizationdirection in order toderive the polarizationmap (image preparedby Sebastian Jester).

Figure 9 (left):LM-band spectra ofIRS 7 with K ~ 6.7, IRS 16 NW with K ~ 9.5, and IRS 33 N with K ~10.5 obtain-ed with the new lowresolution prism.


TTHE ESO VLT SCHEME ALLOWS

astronomers to submit visitormode or service mode observa-tion programmes. In visitormode, the astronomer is pres-

ent at the telescope and can adapt the pro-gramme to specific requirements at the timeof observation. In service mode, the observa-tion details and constraints are submitted toESO beforehand, and the observations arescheduled and carried out by ESO staff.

Service mode observing was conceived byESO in the early days of the planning of VLToperations as a key component in optimizingthe scientific return and the operational effi-ciency (cf. Comerón et al. 2003). The VLTIscience operations scheme profits enormous-ly from the experience gained during servicemode observations at the single UTs sinceApril 1999 and the implemented infrastruc-ture. It follows and is fully integrated into theregular VLT operations scheme from the ini-tial preparation of the proposal to the deliv-ery of the data.

Table 1 shows the numbers of scheduledprogrammes using VLTI instruments (so farMIDI only) for the ESO observing periodsP73 (observations from April to September2004), P74 (October 2004 to March 2005),and P75 (April to September 2005). The firstregular VLTI observing period (P73) hasbeen concluded with a completion rate of80% of the scheduled service mode ob-serving blocks (OBs). These regular VLTIobserving periods were preceded by commis-sioning and shared-risk science observationsusing the K-band commissioning instrumentVINCI, as well as by observations withinMIDI and AMBER science demonstration

and guaranteed time programmes. Theseobservations greatly helped to establish theVLTI-specific aspects of the regular opera-tions scheme as well as to test the overall dataflow system for the VLTI instruments (cf.Ballester et al. 2004).

Also shown in Table 1 are the distributionsof the regular VLTI programmes scheduledso far among programme types, observ-ing modes, as well as scientific categories (B: Galaxies and galactic nuclei; C: Inter-stellar medium, star and planet forma-tion; D: Stellar evolution). Evidently, by far most of these programmes are normal pro-grammes. Aclear majority of programmes areexecuted in service mode, which is partlycaused by the need to combine different base-line configurations (see below). There is anincreasing number of scheduled programmeswithin the scientific category B, while themajority of the VLTI programmes are so farabout equally distributed among scientificcategories C and D. On the subject of this ar-ticle, see also the proceedings of the 2002 Les Houches EuroWinter School “Observ-ing with the Very Large Telescope Interfero-meter”, edited by G. Perrin and F. Malbet. Arecent description of the technical status ofthe VLTI can be found in Glindemann et al. (2004) and references therein. In the fol-lowing, we discuss specific aspects regard-ing VLTI observations as currently offered to the community.

WHY USE THE VLTI FOR ASTROPHYSICS?Modern astronomical observatories have suc-ceeded in measuring the flux density acrossthe electromagnetic spectrum for many cos-mic objects, at least for the brightest objects

in each important class. While one furtherdesires to obtain the flux densities of faint-er and more distant targets, another obviousgoal is to map out each object’s intensitystructure, as a function of wavelength, in asmuch detail as possible (cf. Rees 2000).

For optical and infrared wavelengths, sev-eral interferometric facilities have been oper-ated for this purpose1. These facilities, how-ever, have so far often been limited to thebrightest stars owing to their small collectingareas, and have often not been very easy touse for any astronomer. With the constructionof the Keck and VLT Interferometers includ-ing 8–10 m class telescopes, optical/infra-red interferometry is now also feasi-ble for relatively faint sources. This, for in-stance, promptly enabled the first near- and mid-infrared long-baseline interferometricmeasurements of active galactic nuclei. Inaddition, the VLTI instruments provide un-precedented spectro-interferometric capabil-ities. Finally, as the science operations of theVLT Interferometer were integrated into theregular science operations scheme of the in-struments at the single 8 m telescopes, VLTIobservations can now be relatively easily per-formed by any astronomer.

The range of astrophysical topics that canbe addressed by the VLTI can be seen by look-ing at the science that has already beenachieved using optical/infrared interferom-eters. General results were reviewed by, forinstance, Quirrenbach (2001), Baldwin &Hanniff (2002), and Monnier (2003). Firstscientific results that emerged from the VLTI

THE ESO VLT INTERFEROMETER (VLTI) HAS BEEN IN OPERATION SINCE ACHIEVING FIRST FRINGES IN MARCH

2001. A BROAD SPECTRUM OF ACTIVITIES HAS BEEN COVERED RANGING FROM COMMISSIONING, SHARED-RISK

SCIENCE OBSERVATIONS, SCIENCE DEMONSTRATION AND GUARANTEED TIME OBSERVATIONS, TO REGULAR

OBSERVATION PROGRAMMES IN SERVICE AND VISITOR MODE. THE VLTI OPERATIONS SCHEME IS FULLY INTE-GRATED INTO THE WELL ESTABLISHED REGULAR OPERATIONS SCHEME OF ALL VLT INSTRUMENTS. IN PARTICU-LAR, THE SAME KIND AND LEVEL OF SERVICE AND SUPPORT IS OFFERED TO USERS OF VLTI INSTRUMENTS AS

TO USERS OF ANY INSTRUMENT AT THE SINGLE UTS ON PARANAL. THEREBY, THE VLTI HAS BECOME WORLD-WIDE THE FIRST GENERAL-USER OPTICAL/INFRARED INTERFEROMETRIC FACILITY OFFERED WITH THIS KIND OF

SERVICE TO THE ASTRONOMICAL COMMUNITY.

MMARKUSARKUS WWITTKOWSKIITTKOWSKI , F, FERNANDOERNANDO CCOMERÓNOMERÓN, A, ANDREASNDREAS GGLINDEMANNLINDEMANN, C, CHRISTIANHRISTIAN HHUMMELUMMEL,,SSÉBASTIENÉBASTIEN MMORELOREL , I, ISABELLESABELLE PPERCHERONERCHERON, M, MONIKAONIKA PPETRETR-G-GOTZENSOTZENS, M, MARKUSARKUS SSCHÖLLERCHÖLLER


OOBSERBSERVINGVING WITHWITH THETHE ESO ESO VLVLT IT INTERFEROMETERNTERFEROMETER

1 See the overview provided by Peter Lawson athttp://olbin.jpl.nasa.gov


were reviewed by Richichi & Paresce (2003),and more recent astrophysical results fromthe VLTI are summarized by Wittkowski etal. (this issue, page 36).

TERMINOLOGY RELATED

TO INTERFEROMETRIC OBSERVATIONS

The principle of an observation of a celes-tial light source with an interferometer corre-sponds to the classical experiment by ThomasYoung in 1803 that first showed the wave na-ture of light. Details on the principles of opti-cal interferometry in astronomy can be foundas well in the articles mentioned above (Quir-renbach 2001; Baldwin & Haniff 2002: Mon-nier 2003). Also, the textbook-style pro-ceedings “Principles of Long Baseline Stel-lar Interferometry” of the 1999 MichelsonSummer School in Pasadena, edited by PeterR. Lawson, provide an excellent introduc-tory overview of the technique and appli-cation of optical interferometry (available athttp://olbin.jpl.nasa.gov/iss1999/coursenote.html).

The primary observables of an interferom-eter are the amplitude and phase of theobserved interference pattern, also called“fringe pattern” or just “fringes”. These quan-tities, when normalized, are also referred to as amplitude and phase of the complex“visibility”. The object intensity distribution and the complex visibility are related by aFourier transform. The spatial frequenciesare often denoted u and v, and the Fourierplane is also called the “uv-plane”. The totalflux of the source within the field of view mul-tiplied with the visibility amplitude for agiven experiment is called the “correlatedflux”, corresponding to the “correlated mag-nitude”.

The theoretical visibility amplitude of apoint source is unity for any wavelength and baseline. A source is called “unresolved”(with a given instrumental setup) if its angu-lar diameter is so small that the visibility am-plitude with the given instrumental setup isconsistent with unity within the precision ofthe measurement.

In practice, the measured visibility ampli-tude of a point source would be less than uni-ty due to losses caused by atmospheric andinstrumental perturbations. The real visibili-

ty amplitude of a point source as it would be obtained in an observation is called the“transfer function”. To measure the transferfunction, stars of known, and ideally as smallas possible angular diameter are observed,and the ratio of theoretically expected andmeasured visibility amplitude is derived.This experimental transfer function willchange due to changes of atmospheric andinstrumental conditions on time scales ofminutes to years.

The phase of the observed fringe patternis disturbed by unknown atmospheric pertur-bations of the refractive index above eachtelescope on different time scales down totypically a few milliseconds at optical wave-lengths, the ”coherence time” of the atmos-phere. This means that the observed fringepattern constantly moves and that the integra-tion time that is used to record the interfero-metric fringes can not be much longer thanthis time scale. To overcome this limitation,a fringe-tracking instrument, or “fringe track-er”, can be used that monitors and corrects at high frequency the phase variations of theobserved fringe pattern in order to stabilizeit. Then, a scientific instrument can record thestabilized fringe pattern with a much longerintegration time in order to improve the pre-cision and to observe fainter targets.

Despite the use of a fringe tracker, the un-known differences of the atmospheric pertur-bations above each telescope cause the mea-sured visibility phase not to correspond to theobject phase of the scientific target. There aretwo ways to overcome this limitation. (1) The phase of the complex triple product(also called the “closure phase”), formed bymultiplication of three complex visibilitiescorresponding to three baselines forming atriangle, is not corrupted by the phase noisecaused by atmospheric turbulence, and thusgives directly the object closure phase of thescientific target. (2) If two close targets are

observed through the same atmospheric pat-tern at the same time, and the object phase ofone of the targets is known, the object phaseof the unknown target can be derived relativeto that of the known target. This informationalso includes the angular distance betweenthe (photocentres of the) two targets, and canbe used for astrometry. Also, if one target issimultaneously observed at two sufficientlyclose wavelength bands, the difference of theobject phase at these bandpasses (of the sametarget) can be derived.

OVERVIEW OF THE VLTIThe VLTI comprises four fixed 8 m Unit Tele-scopes (UTs) and four movable 1.8 m Auxil-iary Telescopes (ATs), two of which alreadyarrived on Paranal. The ATs can be positionedon thirty different stations. Figures 1–3 showviews of the VLTI platform. Two scienceinstruments, a near-infrared (AMBER) and a mid-infrared (MIDI) beam-combination in-strument, are in operation.

For period 75, the VLTI was offered withall four UTs, equipped with the MACAOCoudé adaptive optics systems, and the mid-infrared interferometric instrument MIDI.During the year 2005 this configuration willbe extended significantly. First fringes be-tween the first two ATs were achieved on 3 February 2005. This makes it possible touse the VLTI from now on nearly every nightwith competitive telescopes for further com-missioning and technical tasks, as well as forscience operations. It is foreseen to start thescience operations with ATs using a smallsubset of AT configurations, and to expandthe number of offered configurations as ourexperience is growing. For period 76, i. e.observations from October 2005, the MIDIinstrument is offered with all UTs, as well aswith several AT baselines. The near-infraredclosure-phase instrument AMBER is beingcommissioned, has already obtained its first

Period Total Type Mode CategoryNormal GTO DDT s v B C D

73 23 18 4 1 15 8 1 12 10

74 19 14 4 1 13 6 2 8 9

75 24 20 4 – 18 6 3 12 9

Total 66 52 12 2 46 20 6 32 28

Table 1: Scheduled VLTI (MIDI) programmes in ESOobserving periods P73–75. Listed are the totalnumbers of scheduled runs, as well as the distribu-tion of programme types (normal, guaranteed timeobservations (GTO), director discretionary time(DDT) programmes), of programme modes (servicemode ‘s’ or visitor mode ‘v’), and of scientific cate-gory (‘B’: Galaxies and galactic nuclei; ‘C’: Inter-stellar medium, star and planet formation; ‘D’: Stel-lar evolution).

1

2

34

23

Unit Telescopes (1– 4)

Auxiliary TelescopeStations

AT Rail Track

Delay Line Tunnel

Beam CombinationLaboratory

North

Circle with200 m diameter

Figure 1: View of theParanal platform withthe position of the four fixed 8 m VLT UnitTelescopes (UTs) andthe 30 positions for themovable 1.8 m Auxil-iary Telescopes (ATs).Compare with the aerialview in Figure 2 and the recent photographin Figure 3.


scientific data during the last quarter of 2004within science demonstration and guaranteedtime programmes, and is offered for regularobservations with the UTs from period 76(observations from October 2005).

Additional improvements will arise fromthe infrared imaging sensor (IRIS), installedin the VLTI laboratory in early January, which will stabilize the image of the observedobjects in the focus of the instruments. Thefringe tracker FINITO will hold the interfer-ometric fringes in such a way that the scienceinstruments can integrate them for muchlonger times than otherwise prescribed byatmospheric turbulence. Finally, the phasereference and micro arcsecond astrometry fa-cility (PRIMA) will allow astrometric meas-urements with high accuracy and to directlyretrieve spatial phase information of scientif-ic targets.

THE MIDI INSTRUMENT

MIDI is the mid-infrared instrument of theVLTI. It combines two beams (either fromtwo UTs or from two ATs) to provide visibil-ity amplitudes. The light is dispersed afterbeam combination with a spectral resolu-tion of either R ~ 30 (prism mode) or R ~ 230(grism mode). The limiting magnitude for un-resolved sources, i. e. the limiting correlatedmagnitude, is currently N = 3.25 (2 Jy) forservice mode observations using the UTs andthe prism mode. For visitor mode observa-tions, when the investigator is present at thetelescope and can decide whether the ac-quisition image and achieved beam overlap fulfills expectations, ESO accepts expec-ted correlated magnitudes down to N = 4(1 Jy). More technical information regarding the MIDI instrument can be found at http://www.eso.org/instruments/midi/.

THE AMBER INSTRUMENT

AMBER is the near-infrared phase-closureinstrument for the VLTI. The instrument hasbeen designed to combine simultaneouslythree beams, coming from a triangle of tele-

of calibration stars preserves objects whichhave already been studied. Hence, more andmore detailed knowledge of these calibrationsources will be rapidly acquired (cf. Ballesteret al. 2004). For work related to diameter esti-mates of calibration stars, see also the last sec-tion in Wittkowski et al. (this issue) and ref-erences given there.

Proposals for observations using VLTI in-struments are prepared and submitted us-ing the standard ESO tool as for any VLTobservation (see http://www.eso.org/observ-ing/proposals/). For VLTI observations, theproposal includes an additional interferomet-ric table that lists the expected angular sizesof the proposed targets and their expected vis-ibility values and correlated magnitudes (asfor instance obtained with VisCalc).

The actual observations are prepared withthe standard phase 2 proposal preparation(P2PP) tool. A detailed description of obser-vation preparation, observation tools, phase2 and OB preparation can be obtained fromthe pages of the ESO User Support Depart-ment (USD) at http://www.eso.org/org/dmd/usg/. Just as with service mode support forany other VLT instrument, VLTI observerscan obtain assistance from astronomers at theUSD specialized in interferometry.

SEQUENCES OF OBSERVATIONS

In order to obtain sufficient accuracy and pre-cision of the object visibility values, observ-ing sequences with alternating observationsof scientific targets and interferometric cali-bration stars are performed. The data takenon all calibration stars are public once theyarrive in the ESO archive. Hence, each inves-tigator can make use of the information basedon all measured calibration stars.

The scientific goal of an interferomet-ric observation campaign can often only bereached if visibility measurements at a rangeof different points of the uv-plane are com-bined. This can be achieved by combiningdifferent ground baselines and by making useof Earth’s rotation.

scope stations, in order to obtain closurephases. The combined beams are spectrallydispersed with resolutions of R ~ 30 (low res-olution, LR), R ~ 1500 (medium resolution,MR), or R ~ 10 000 (high resolution, HR).This means that each instantaneous AMBERmeasurement provides three sets (a set in-cludes a number of spectral channels) of visi-bility amplitudes, for the three baselines com-prising the triangle, as well as one set of triple products (including triple amplitudeand closure phase). More technical informa-tion regarding the AMBER instrument isavailable at http://www.eso.org/instruments/amber/.

USER SUPPORT AND PREPARATION TOOLS

FOR THE VLTIIn the same way as for other VLT instruments,all relevant information for the use of VLTIinstruments is provided by ESO throughdifferent standard documents and via thestandard ESO webpages, including the callfor proposals, the instrument and templatemanuals, as well as the webpages with gen-eral and instrument-specific proposal andobservation preparation instructions.

In order to assess the feasibility of aplanned observation with MIDI or AMBER,it is mandatory to estimate the visibility val-ues for the expected intensity distribution ofthe science target and chosen VLTI configu-ration. The interactive tool provided anddeveloped by ESO to obtain such visibilityestimates is VisCalc (Visibility Calculator).A more detailed description of this tool in-cluding examples can be found in the recentarticle by Ballester et al. (2004).

The second tool provided and developedby ESO to support VLTI observation prepa-ration is CalVin, which may be used to select,for each science target, a calibration star (cf. Ballester et al. 2004). Based on differentuser-defined criteria, CalVin selects suitablecalibrators from an underlying list of calibra-tors. The strategy to preferably select calibra-tion stars from the limited underlying lists

Figure 2: An aerial viewof the Paranal platformas of December 1999,still in construction,with the four 8 m UTsand several founda-tions of the 30 stationsfor the 1.8 m ATs.

al spectral channels and (projected) baselineconfigurations, are combined, it may alreadybe possible to directly reconstruct the imageof a simple object intensity distribution, forexample of a binary star. The first optical im-ages reconstructed from such sparsely sam-pled visibilities and closure phases, obtainedwith the “Cambridge Optical Aperture Syn-thesis Telescope (COAST)” and the “NavyPrototype Optical Interferometer (NPOI)”,were presented by Baldwin et al. (1996) andBenson et al. (1997), respectively. However,in the majority of cases so far, and probablyas well in the near future, the interpretationof optical/infrared calibrated visibility valuesand closure phases from long-baseline inter-ferometry is accomplished by comparison tomodels. A model-predicted object intensitydistribution is used to compute the correspon-ding synthetic visibility values and closurephases for the employed baseline configura-tions and spectral channels, and these valuesare compared to the measured ones in orderto find agreement or disagreement. Usually,model parameters are adjusted in order to findthe best possible agreement with the mea-sured data.

ACKNOWLEDGEMENTSThe integration of VLTI observations into the reg-ular ESO science operations scheme has involveda lot of people inside and outside ESO, too manyto name here. We warmly acknowledge this collec-tive effort.

REFERENCESBaldwin, J. E., Beckett, M. G., Boysen, R. C. et

al. 1996, A&A, 306, L 13Baldwin, J. E. & Haniff, C. A. 2002, Phil. Trans.

R. Soc. London, Ser. A, 360, 969Ballester, P., Percheron, I., Sabet, C. et al. 2004,

The Messenger, 116, 4Benson, J. A., Hutter, D. J., Elias, N. M., II, et al.

1997, AJ, 114, 1221Comerón, F., Romaniello, M., Breysacher, J.,

Silva, D. & Mathys, G. 2003, The Messenger,113, 32

Glindemann, A., Albertsen, M., Andolfato, L. et al. 2004, Proc. SPIE, 5491, 447

Monnier, J. D. 2003, Reports on Progress in Phys-ics, 66, 789

Quirrenbach, A. 2001, ARA&A, 39, 353Rees, M. J. 2000, Proc. of IAU Symposium 205,

p1Richichi, A. & Paresce, F. 2003, The Messenger,

114, 26

17© ESO – March 2005Wittkowski M. et al., Observing with the ESO VLT Interferometer

The sky-projected baseline length and angle,as well as the zenith distance, are uniquelydefined for a given target and ground base-line configuration by the hour angle, or thelocal sidereal time, at which the observationis executed. The local sidereal time (LST) atwhich the observation shall be executed canbe inserted into each observation block (OB).By this, observations at different sky-pro-jected baseline lengths and angles can beplanned, while the individual OBs are exe-cuted as stand-alone entities without the needof linked observations. However, priority isgiven to completing all observations on a sci-entific target once they have started. Thepreparation and planning of such observationsequences is supported by the visibility cal-culator VisCalc (see above).

THE CONSTRAINT SETS AND SCHEDULING

OF VLTI OBSERVATIONS

The required ground baseline configuration,as well as the local sidereal time at time ofexecution (see above) are constraints that arespecific to VLTI observations and do not ex-ist for observations with instruments at thesingle UTs. Moreover, the performance ofVLTI-specific subsystems such as the adap-tive optics systems (MACAO) for VLTI, orthe fringe tracker have to be monitored inaddition.

On the other hand, other constraints, as for example the requirements for the regularseeing condition or for the lunar illuminationmay be less important for VLTI observationsthan for classical VLT instruments such asISAAC or the FORSes. In general, the sched-uling of VLTI observations is complicated bythe additional constraints on baseline con-figuration and LST. As a result, one tries toavoid additional stringent constraints in orderto enable a smooth and efficient science oper-ation of the VLTI.

QUALITY CONTROL OF VLTI DATA

Since P73, the MIDI data pass through theentire data flow operations systems in thesame way as the data of any other VLT instru-ment (cf. Ballester et al. 2004). The raw filesobtained during service mode observationsare processed for quality control purposesusing the data reduction system developed byESO, the algorithms of which have beenprovided by the MIDI consortium. Eachprocessed OB is associated with a FITS fileproduct containing quality control parame-ters and results, such as the visibility valueand the instrumental transfer function for dataof known calibration stars, or a calibrated vis-ibility value for a scientific target. For P73,80% of the OBs received by the pipelineresulted in data products. For the start of theAMBER science operations, a similar dataflow operations system will be available forAMBER data.

REDUCTION OF VLTI DATA

The ESO pipeline data reduction as describedabove was developed for quality control pur-poses and optimized to process most of thedata obtained (with a large variety of flux lev-els, visibility amplitudes, etc.) in an automat-ic way with acceptable accuracy and preci-sion. A more detailed reduction of data froma specific programme may be necessary forscientific purposes which usually have morestringent requirements for precision andaccuracy.

Two data reduction packages are current-ly available for the scientific reduction ofMIDI data, available from MIDI consortiummembers of the University of Leiden and theMax-Planck-Institute for Astronomy in Hei-delberg (a merged version can be found athttp://www.strw.leidenuniv.nl/~nevec/MIDI/index.html; also see http://www.mpia-hd.mpg.de/MIDISOFT/). A third one from theObservatory of Meudon is about to be re-leased to the astronomical community. Thesepackages have some technical differences,the details of which can be found in their doc-umentation. For a given specific purpose, oneof these packages may be more suitable thananother. Detailed technical information onMIDI data reduction is also being collectedand compiled by Christian Hummel, andprovided to the user community on an “as is”basis at http://www.sc.eso.org/~chummel/midi/midi.html.

So far, one package exists for the reduc-tion of AMBER data, available from consor-tium member Astrophysical Lab of the Obser-vatory of Grenoble. The AMBER packagehas not yet been released to the public.

MIDI data reduction will usually result in one calibrated visibility amplitude for each spectral channel and observation. ForAMBER, the resulting processed and cali-brated data will usually include three cali-brated visibility amplitudes, correspondingto the three baselines of the triangle, and onecalibrated triple product (triple amplitude andclosure phase) for each spectral channel andobservation.

If such interferometric data, obtained witha three station array such as the VLTI with the AMBER instrument and taken in sever-

Figure 3: A recent photograph takenon the VLTI platform in January 2005by Bertrand Koehler. The first twoATs, located at station E0 and besideG0, are shown in the foreground, as well as UTs 3 and 4 (MELIPAL andYEPUN) in the background. Thebuilding that can be seen in betweenthe ATs is located on top of theinterferometric lab, where the beamscoming from the UTs or ATs are com-bined. The beams are sent throughan underground light-duct system to the interferometric lab. Courtesy B. Koehler.


IIN THE EARLY DAYS OF THE 3.6 M THE

delivered image quality was not opti-mal. Image quality analysis was donealmost always at the zenith using the“pupil plate”, a photographic method

with defocused images. The image qualitywas not verified systematically at positionsfar from the zenith. Coma aberration was cor-rected by a time consuming mechanical andmanual re-collimation of the M2 tilt. Otheraberrations were present but found to be un-stable and with unclear origins at that time.Mechanical analysis of the telescope flexureperformed during the 1980s showed incorrectbehavior of the top unit with hysteresis pat-tern (Figure 1).

Observations at the Cassegrain focus us-ing CCD detectors have brought to light fur-ther limitations on image quality. The use ofa seeing monitor confirmed large differencesbetween the atmospheric seeing and that ofthe 3.6 m. On average, by the early 1990s, the3.6 m image quality was above 1 arcsec.

Using a portable wavefront sensor (Shack-Hartmann) with a CCD detector we started asystematic campaign to investigate the imagequality. The first results obtained in 1991 con-firmed the degradation of the image qualityas a function of Zenith distance (Gilliotte,2001).

Several concerns were identified as: (1)Thermal effects were clearly important to the image quality stability. Defocus, spheri-cal aberration, tilt (image stability) and alsoastigmatism terms contributed, to the totalimage quality up to 1 arcsec when the tele-scope dome area was warmer than outside;(2) Coma, triangular and astigmatism aberra-tions changed, as a function of telescope posi-tion. Values up to 0.9, 0.5 and 0.6 arcsec wereseen; (3) A constant value of spherical aber-ration was measured (around 0.5 arcsec).

The graph in Figure 1, prepared in 1991 bya mechanical engineer (J. Cheng), illustratesthe different contributions to the total meas-ured displacement of M2 with respect to M1.It was clear that the telescope image qualityneeded to be improved.

THE 3.6 M UPGRADES

In 1995 the goal of the 3.6 m upgrade projectwas to obtain a sub-arcsec image quality of0.9 arcsec over 120 deg solid angle for the

average site seeing. Following detailed stud-ies almost all telescope parts and the domewere subject to this intervention. The domeand mirror seeing were reduced by minimi-zation of heating sources in the dome, dome air cooling, and forced ventilation in front of the mirror. The spherical aberration was cor-rected by lowering the focal plane by 166 mm.

Triangular aberration (up to 0.4 arcsec at60 deg zenith distance) was found to varywith telescope elevation and a modificationof the M1 cell was made. This aberration wasreduced below 0.2 arcsec by means of con-stant force components (springs) on the M1axial astatic levers. A new wavefront sensingmethod was used by means of direct CCDobservations (Curvature Sensing).

During the 1990s, all efforts were ori-ented towards the main mirror cell and thethermal environment problems (cf. Gilliotte,2001). By 2000 the image quality of the tele-scope was still limited by the coma insta-bilities (up to 0.6 arcsec). Close to the zenith the telescope delivered an image quality of0.7 arcsec at best.

New instruments (HARPS) and the in-creasing demand for high spatial resolutionobservations (0.16 arcsec for EFOSC2) pro-vided the incentive for further image quali-ty improvement. Upon completion of the M1cell upgrade a new project to improve the M2unit started in 2001.

THE M2 UPGRADE

The following goals defined the require-ments: (1) Remote collimation of M2 forcoma correction; (2) Coma must be betterthan 0.1 arcsec after applying the collimation

correction; (3) Minimization of residual tele-scope flexures and related hysteresis; (4) Cor-rection of the M2 focus instabilities.

The Coudé M2 unit support and the NTTM2 collimation concept were used as startingpoints. A new complete top ring was built. A pantograph design was used for the M2support allowing the M2 to tilt around thecenter of curvature by means of an x, y trans-lation table.

The new unit was installed in August 2004and immediately gave an improvement inimage quality and ease of focusing operation.However hysteresis on the coma variationwas still observed, preventing a precise col-limation correction. The M2 cell was then the last untouched part of the 3.6 m and pos-sibly responsible for the residual behaviour.A complete maintenance of the cell includ-ing a realuminization of M2 was performedin November 2004. Additional unexpectedsources of instability in coma and astigma-tism were identified and corrected.

OPTICAL QUALITY RESULTS

The results obtained after the November 2004intervention demonstrate that the 3.6 m tele-scope finally delivers an excellent imagequality. Figure 2 illustrates the coma vari-ation with respect to the S-N axis of thetelescope. This axis has the strongest comaaberration. The residual coma hysteresis dis-appeared completely.

Figure 3 shows the total aberration varia-tion on the full sequence S-N-W-E withoutM2 collimation correction. The final imagequality in terms of classical optical aberra-tions is less than 0.4 arcsec, with a very small

A SUMMARY OF THE 3.6 M IMAGE QUALITY IMPROVEMENT IS PRESENTED WITH THE LATEST M2 UPGRADE

AND ITS RESULTS.

AALAINLAIN GGILLIOTTEILLIOTTE , G, GERARDOERARDO IIHLEHLE , G, GASPASPAREARE LLOO CCURURTOTO


IIMPROVEMENTSMPROVEMENTS AATT THETHE 3.6 3.6 MM TTELESCOPEELESCOPE

Figure 1: M1/M2 rela-tive displacement withDeclination.


amount of residual coma (< 0.2 arcsec) with-out collimation correction (< 0.1 arcsec withcollimation correction). The main residualterm is the astigmatism (as expected), contri-buting 0.35 arcsec when pointing far to thenorth.

SCIENCE RESULTS

During the last months the improvement ofthe image quality of the telescope has beenclearly noticed in EFOSC2, CES and HARPSobservations.

The EFOSC2 seeing is measured in termsof image size and has been logged for sever-al years. It includes telescope and instrumentimage quality and atmospheric seeing. In Fig-ure 4 the evolution of the EFOSC2 seeingversus the corresponding measurements of the seeing monitor is presented. This graphcovers several EFOSC2 runs. Each verticalband corresponds to one observing period.The red squares represent the seeing monitorvalue, while the blue squares correspond tothe EFOSC2 image size in the y direction.During the most recent runs, the EFOSC2measurements were comparable to the seeingmonitor values down to 0.45 arcsec. Beforethe M2 upgrade the seeing as measured inEFOSC2 science images would not reach val-ues lower than 0.7 arcsec, being limited bythe telescope image quality.

HARPS observers have also noticed theimprovement in image quality, gaining up to40% more flux with respect to the pre-M2upgrade times. In Figure 5 the gain in ef-ficiency after the two M2 interventions(August 2004 and November 2004) is illus-trated. Horizontal lines indicate averagevalues.

The benefits of the new M2 unit are alsonoticed from the operational point of view:focusing operations are simpler and fasterand coma aberrations can be corrected by remotely re-collimating the secondarymirror, although this option is only need-ed when observing to the extreme north.Since the November 2004 intervention, coma (< 0.2 arcsec) is independent of telescopepointing direction.

We look forward to many years of smoothoperation and excellent image quality withthe 3.6 m telescope.

ACKNOWLEDGMENTSWe are grateful to Jaime Alonso and Ismo Kastinenfor their fundamental contribution on the electron-ic and software part of the M2 upgrade project. TheLa Silla Mechanical team has been working hardall the time on the 3.6 m project and the success ofthe complete project is certainly in a large part dueto them. Ueli Weilenman and Stephane Guisardalso made major contributions through this pro-ject, leading the first M1 upgrade and in charge of the optical analysis respectively. We also thankFrancesco Pepe for supplying the illustration of theHARPS efficiency.

REFERENCESGilliotte, A. 2001, The Messenger, 103, 2

Figure 2: ComaVariation with ZenithDistance.

Figure 3: Image qualityimprovement: OpticalAberration variationwith Zenith Distance.

Figure 4: EFOSC2 see-ing history (blue andgreen) compared withseeing monitor (red).

Figure 5: HARPSEfficiency variation with observing run.


EEVERY SIX MONTHS ABOUT 900scientific observing proposalsare written to make use of ESOtelescopes. Proposals are eval-uated by an external Observing

Programmes Committee (OPC), which rec-ommends the allocation of telescope time viaa ranked list of proposals (see Figure 1). Thegoal of the Time allocation Tool (TaToo) is to schedule the telescopes in the most opti-mal and reliable manner possible, taking intoconsideration the full set of constraints ofeach OPC recommended observing program.TaToo is not intended to be a fully automat-ed “black-box” program, but a user friendly,interactive, semi-automated tool used byESO’s Visiting Astronomers Section (VISAS)to generate and maintain the long-term sched-uling of ESO telescopes.

Today, after successfully scheduling thelast two observing semesters with TaToo,we must take a step aside to pay tribute to the fomer Head of VISAS, Dr. JacquesBreysacher, who scheduled ESO’s telescopesfor almost 30 years (see Figure 2). Dr.Breysacher was the initiator and the strongestsupporter of the TaToo project and perhapsthe only one who can fully appreciate theintricacies of an automated scheduler forESO telescopes. His experience and strongsense for practical solutions were fundamen-tal during the development of TaToo.

OUR APPROACH TO THE

SCHEDULING TECHNOLOGY

The challenge to develop a software tool witha high production quality has forced us tomake a very careful choice of the underlyingscheduling technology. The reliability of thistool and the quality of the schedule producedare of paramount importance to the observa-tory. A schedule solution that secures optimalobserving conditions for each recommendedprogram and maximizes the number of pro-grams on the telescopes contributes decisive-ly to the effective usage of ESO telescopes,boosting the scientific return of the observa-tory.

At the beginning of TaToo’s design, in 2003,we did an extensive evaluation of the existenttelescope scheduling systems. Among thedifferent systems were 1) Spike by Johnstonand Miller (1996), 2) the system by Grim et al. (2002) based on a genetic algorithmsscheduler, 3) the “Just-in case” telescopescheduling algorithm developed by Drum-mond et al. (1994), etc. Of all of these, onlySpike was an established telescope sched-uling system that due to its modular constraintsatisfaction solver was flexible enough andcould potentially be adapted for use at ESO.An attempt to adapt Spike at ESO is describedby Giannone et al. (2000). However, thedesign of the Spike scheduler had been donein the early 1990s, at a time when constraintprogramming was still at its early stages ofdevelopment. Contemporary constraint pro-gramming systems include a large number ofvery powerful search and constraint propaga-tion techniques that offer more effectivescheduling, see, e.g., Baptiste et al. (2001).

This conclusion, as well as a careful studyof the available open-source/commercial op-timization and scheduling technology, allow-ed us to define our approach to the develop-ment of TaToo as follow: “Select a modernand real-life proven scheduling technology,and focus efforts on the interface with theESO scheduling problem”. It was quite clearfrom the beginning of the project that devel-oping a new scheduling technology fromscratch was beyond the scope and budget ofthe project. Instead, most of the one year wehad to complete the project would have to bespent translating the ESO scheduling prob-lem to the language of a well-establishedscheduling technology.

During our search for the best schedulingtechnology on the market we analyzed sys-tems based on: – Genetic algorithms: i2 (2002).– Linear, quadratic and integer optimization

systems: Optimization Solutions Library– IBM, see COIN (2002); Xpress – DashOptimization (2002), CPLEX – ILOG(2002).

TATOO IS ESO’S NEW TIME ALLOCATION TOOL. THIS SOFTWARE SCHEDULER IS A COMBINATION OF A USER-FRIENDLY GRAPHICAL USER INTERFACE AND AN INTELLIGENT CONSTRAINT-PROGRAMMING ENGINE FINE-TUNED

TO ESO’S SCHEDULING PROBLEM. TATOO IS ABLE TO PRODUCE A HIGH QUALITY AND RELIABLE SCHEDULE

TAKING INTO CONSIDERATION ALL CONSTRAINTS OF THE RECOMMENDED PROGRAMS FOR ALL TELESCOPES IN

ABOUT 15 MINUTES. THIS PERFORMANCE ALLOWS SCHEDULERS AT ESO-VISAS TO SIMULATE AND EVALUATE

DIFFERENT SCENARIOS, OPTIMIZE THE SCHEDULING OF ENGINEERING ACTIVITIES AT THE OBSERVATORIES, AND

IN THE END CONSTRUCT THE MOST SCIENCE EFFICIENT SCHEDULE POSSIBLE.

JJOÃOOÃO AALLVESVES,, EUROPEAN SOUTHERN OBSERVATORY

TTELESCOPEELESCOPE TT IMEIME AALLOCALLOCATIONTION TTOOLOOL

Figure 1: The workflow of the long-term schedulingprocess at ESO: For each 6-month semester a large number (currently about 900) of observingproposals are submitted to ESO. The independentObserving Programmes Committee (OPC) evalu-ates all proposals and recommends time allocationby creating a ranked list where the proposals are ordered according to their scientific merit. Thetechnical feasibility of the proposals is checkedduring the ESO’s technical evaluation. The final list(OPC ranked) is then used by VISAS as input to the long-term scheduling. The final schedule isstored in an ESO database and published in weband spreadsheet forms.

Scientific Evaluation ofProposals by OPC

ESO technicalevaluation

Export toESO Web

Export toSpread-sheet

Submission ofObservingProposals

Long-termScheduling

Final schedule

Ranked list

Final list

Figure 2: First ESO tele-scope schedule com-puted by Dr. JacquesBreysacher, for Period22 (1978). The know-how accumulated dur-ing almost 30 years

– Constraint programming: CHIP V5 –COSYTEC (2002); clp(fd) – Diaz (2002)and IC-Parc (2002), open source; Solver/Scheduler – ILOG (2002); Mozart /Oz(2003) – DFKI, open source; Koalog Con-straint Solver (2003).

In order to experiment with the different algo-rithms and modeling strategies, and to eval-uate performance, we developed a prototypeof TaToo. Finally, after comparison betweena complete set of results, we selected the com-bination Solver/Scheduler of the French com-pany ILOG. The Solver is a library for con-straint programming while the Scheduler setsan additional abstraction layer over the Sol-ver that simplifies and optimizes the model-ing through notions like activities, resources,reservoirs, states, precedences, etc. These twolibraries are being used by many organiza-tions like Deutsche Bahn, SAP, Lufthansa,Daimler Chrysler, Deutsche Telekom, BMW,Nippon Steel, NFL, IBM, Metro de Madrid,etc. – see Connection (2003).

The software package of ILOG containsan Integrated Development Environment(IDE) with debugging functions that we usedextensively during the development of sched-uling models and defining optimal searchstrategies (Figure 3).

TATOO’S ARCHITECTURE

The architecture TaToo is shown in Figure 4.The entire scheduling and control logic ishosted on the Scheduling Server. The data arestored in two databases on the Database Serv-er(s). The clients access the system througha (fat client) graphical user interface (GUI).

Each observing program sets a range of re-quirements and conditions on the scheduling.The Control Logic reads them from the Ob-serving Proposals database and transfersthem to the Scheduling Engine. There, prop-er constraints are generated and sent to theSolver/Scheduler together with the corres-ponding constraint Models. The schedulingresults are written back to the Operation-al Data database. The system operator hasaccess to all relevant data and control overthe entire scheduling process via the GUIClient (Figure 5 and Figure 6). The Modelsare written in Optimization ProgrammingLanguage (OPL), the Control Logic and theScheduling Engine in Perl, the GUI Client inJava. The libraries Solver/Scheduler are pre-compiled (written by ILOG in C++).

HOW DOES TATOO SCHEDULE?There are two modes of scientific observa-tions at ESO telescopes: the Visitor mode(VM) and the Service mode (SM). VM is theclassical mode of observations in which theobservations are executed by an astronomerfrom the proposing team that is physicallypresent at the telescope. SM observations, onthe other hand, are performed by the ESOobservatory staff. VM observations consist ofruns; a run may be additionally divided into

Figure 3: The OPL Stu-dio of ILOG used forthe development of theConstraint Program-ming models of TaToo.The panels show (from right to the left)the source code of the OPL (OptimizationProgramming Lan-guage) model; the solu-tion search tree; theearliest/latest timespans of each variable;the data structure of the model. The lowerpanel shows theprogress of the optimi-zation.

of scheduling ESO tele-scopes was fundamen-tal in the translation of the ESO schedulingproblem to schedulingtechnology language.



sub-runs. A sub-run is the smallest schedul-able entity and occupies at least half a night(Figure 7). SM observations, on the otherhand, are performed in one-hour observationblocks.

From TaToo’s perspective, the substantialdifference between VM and SM observationsis the search space in which the schedulingtakes place. The VM (sub)runs are scheduledon the time axis, meaning that each scheduledVM run becomes a particular, fixed time spanfor execution. The SM runs are scheduled ina “resources” space. A scheduled SM run isone that has been accepted in the schedule if sufficient resources for its execution areavailable. The observatory staff determineswhen a scheduled SM run will be performedafter considering the meteorological condi-tions, the current states of the queues, and itschance of getting substantially completed bythe end of the observing semester.

TaToo schedules VM runs by proper OPLmodels. The models take into account allparameters important for the run like OPC-ranking, object coordinates, required moonillumination, sub-runs configuration, angulardistance to the moon, critical and avoid-dates,etc. (Figure 8 and Figure 9). These parame-ters are used to generate the correspond-ing constraints of the models. In some cases, e.g., to minimize the number of instrumentchanges, the models themselves define addi-tional constraints at run time. The effectivealgorithms for constraint propagation imple-mented in Solver/Scheduler libraries as wellas the properly selected search strategies in the models lead to very good schedulingperformance. On a 2 GHz single processorcomputer the scheduling of all seven tele-scopes takes less than 15 minutes. In this time≈ 100.000 constraints and 500–1000 sub-runs per telescope are evaluated and (some ofthem) scheduled.

For the scheduling of SM observationsTaToo implements a two-step procedure:

Step 1: On this step TaToo generates theso-called pseudo-VM (PVM) runs. The gen-eration works in the following way. The RAcoordinate of each target of each requestedSM run is used to define a visibility windowwhere the target can potentially be observed.A new PVM run is defined for that target,including the required moon illumination asa constraint. To compensate the different timeresolution of VM (0.5 nights) and SM (1 hour)a procedure fills-up the 0.5 night block inPVM by adding other relevant targets of thesame SM run or of rank-neighboring SMruns.

Depending on the VM/SM time distribu-tion and on the particular SM pressure at eachtelescope, a large number of interchangeablePVMs may be generated. Aspecial procedureanalyzes the configuration of the generatedPVMs and removes the logical symme-tries by generating sets of additional prece-dence constraints. This substantially prunes

ExportedSchedules (web,

spreadsheet)

SchedulingServer Clients

OperationalData

ObservingProposals

DatabaseServer(s)

Control Logic

ManagementConsoleReports &

Statistics

GUI ClientModels

Solver/Scheduler

Scheduling Engine

Figure 4: Architectureof TaToo.

Figure 5: At any pointof the schedulingprocess the graphicaluser interface of TaTooprovides access andcontrol of all relevantfor the scheduling data.

Figure 6: Graphical pre-sentation of the finalschedule in a timetableform. The instrumentsare color-coded. Thepink color denotes timeallocated for SM runs.The panes with tablesbelow the timetableprovide detailed infor-mation about eachscheduled run.

23© ESO – March 2005Alves J., Telescope Time Allocation Tool

the search tree and increases the overallscheduling performance.

Finally, the PVMs are competitivelymixed with the VM (sub)runs by taking intoaccount the OPC ranking list (Figure 10) andare fed to the OPL models for scheduling.

Step 2: During this step of the SM sched-uling TaToo implements an algorithm basedon the ones described in Silva (2001). Thealgorithm uses a RA/MOON/SEE/TRANS(RMST) model and schedules by consump-tion of time resources. The calculation of theavailable time resources is based on statisti-cal data about the weather conditions at theobservatories’ sites and is performed for thetime spans of the PVM runs scheduled dur-ing Step 1.

The described SM scheduling procedureprovides a fair time assignment, especially inthe over-subscribed RA-ranges (see Alves &Lombardi 2004) as it leverages the advan-tages of both the constraint programmingmodels and the RMST model.

FINAL REMARKS

One of the most important characteristics of TaToo is its overall performance. TaToo isable to produce a high quality and reliableschedule taking into consideration all con-straints of the recommended programs for alltelescopes in about 15 minutes. This is cru-cial for a final optimization level where theTaToo operator, an astronomer, can simulateand evaluate different scenarios (e.g., furtherdiffusing oversubscribed RA’s, assessing theimpact of an unpredictable instrument fail-ure, etc.) in more or less real time. These sim-ulations also allow for an optimal long-termscheduling of large engineering time blocks(small engineering time blocks and instru-ment calibrations are automatically sched-uled by TaToo), enabling the ESO schedulersto construct the most science-efficient sched-ule possible.

Finally, users must keep in mind that someprograms, even programs highly ranked bythe OPC, might not fit the schedule due toexhaustion of a particular combination ofobserving conditions (Moon illumination,Seeing, etc.). Typically these cases occurwhen proposals request highly demandedRA’s where competition with Large Pro-grams and higher ranked programs reaches amaximum. While the number of highlyranked programs that do not fit a particular

Figure 8: Illustration of the way the scheduler determines the Earli-est /Latest Interval where a VM run containing two sub-runs may bescheduled. For simplicity, the figureshows only some of the constraintsapplied. In reality many more con-straints such as critical and avoid-dates, linked runs, proper half-nights,scheduling runs of the same PI close together, minimizing of instru-ment setup time, etc. are applied.

Figure 10: TaToo generates the finalranked list by normalizing and merg-ing the lists of all eight OPC sub-pan-els: (1) For each sub-panel the list ofproposals above the cut-off line isnormalized between 0 and the cut-offline. (2) The normalized lists of allsub-panel are merged together. (3) Incase proposals on the final rankedlist overlap (like proposals A1, 3 andA2, 4 on the figure), the proposalsubmitted earlier is given advantageand is ranked higher. (4) Steps 1–3are repeated for the proposals belowthe cut-off line.

Figure 9: On the upper panel TaTooshows the target visibility (number ofobservable hours per night) for eachtarget and the time window (the yel-low box) in which the observation runmay be scheduled. The second and the third panels show the visibili-ty during the first (H1) and the sec-ond (H2) half-nights. The fourth panelillustrates the angular distance ofeach target to the moon. The bluerectangle drawn at 30° shows theminimal allowed angular distance.

Figure 7: The VM run shown hereconsists of 5 sub-runs. The “3H1” arethree first half-nights, followed by “2n” – two whole nights and “1.5n” –1.5 nights starting at the beginning of a night. The required intervals thesub-runs are 3 nights ± 50% be-tween sub-runs “3H1” and “2n” and4 nights ± 50% between sub-runs“2n” and “1.5n”. The diagonally-striped gray areas show the areaswhere sub-runs “2n” and “1.5n” maybe scheduled, provided sub-run“3H1”is on a fixed position. Actually,TaToo tries to find optimal positionsof all three sub-runs simultaneouslyby introducing from- and to-limits ofthe distance constraints.

2...4 n 3..5 n

2 n

time

3H1 1.5 n

cut-offline

0

1

2

3

4

1

2

3

1

2

3

4

5

6

7

sub-panel A1 sub-panel A2 final ranked list...

Erliest / LatestInterval

Visability

t

t

Reserved

t

CloseMoon

t

70%

40%

MoonIllumination

t

Requested

t


THE CURRENT STATUS OF THE ANTENNA

PROCUREMENT

Presently the ALMA antenna procurementprocess is being delayed until further tests ofthe prototype antennas in Socorro NM, USAare finished. These tests involve some astro-nomical measurements, so winter is the mostfavorable time period. Once the tests are fin-ished the results will be evaluated and a deci-sion about the choice of ALMA antenna willbe made. As all should understand, great cau-tion is needed in reaching this decision, sincethe ALMA antennas will be the largest singleinvestment in the project.

ESAC MEMBERS

From Janunary 2005, José Cernicharo has be-come the Spanish member of the EuropeanScience Advisory Committee (ESAC). Hereplaces Rafael Bachiller. The names of theother national members are to be found at theweb site http://www.eso.org/projects/alma/newsletter/almanews2/ESAC.

THE PRESENT STATUS OF THE

ALMA REGIONAL CENTER

The concept of the ALMA Regional Center(ARC) for Europe has been discussed by the European Science Advisory Committee(ESAC) in September 2003. This discussionis summarized in an appendix to the ESACreport. After further discussions within theEuropean ALMA Board, the STC and ESOCouncil, the ESO Council approved a “Callfor Expressions of Interest”, with the requestto submit letters of intent by 31 October 2004.Seven replies have been received. These willbe discussed in a face-to-face meeting at ESOin early 2005 with the groups involved. Thusprogress is being made on the organization ofARCs, and we will provide more details infuture issues of The Messenger.

For those interested in the background, theARC functions are divided into “User Sup-port”, which is funded within the ALMAproject, and “Science Support” which is nota part of the basic ALMA funding plan.Recent accounts of the “User Support” areto be found at the web site: http://www.eso.

org/projects/alma/meetings/gar-sep04/Silva_Community_Garching.pdf

For a description of “Science Support”,see the web site http://www.eso.org/projects/alma/meetings/gar-sep04/Wilson_Commu-nity_Garching.pdf

For other presentations of functions given at the ALMA Community Day, see http://www.eso.org/projects/alma/meetings/gar-sep04/

UPCOMING EVENTS

There will be a workshop entitled “SZ Effectand ALMA” on 7–8 April 2005, at Orsay, inthe Paris area. For further information andregistration, email [email protected]

Planning has been started for a “GlobalALMA Meeting” to be held in Madrid in2006. This will be the first world-wideALMA science meeting since the Washing-ton DC meeting in 1999. The local organiza-tion of the meeting will be headed by RafaelBachiller (OAN), while the scientific organi-zation will be led by the Alma Scientific Advi-sory Committee.

ALMAALMA NNEWSEWSTTOMOM WWILSONILSON, EUROPEAN SOUTHERN OBSERVATORY

schedule is very small (typically a fewprograms per semester), even these could beavoided if proposers find targets in lessdemanded RA’s (see Alves & Lombardi2004).

ACKNOWLEDGEMENTSI would like to thank Dr. Vassil Lolov for the expert-ise, bright ideas, and motivation given to this proj-ect. Dr. Lolov worked on the TaToo project undercontract with ESO. I also would like to thank therest of the VISAS team as well as all ESO col-leagues that helped with suggestions during thedevelopment of TaToo.

REFERENCESAlves, J., Lombardi, M. 2004, The Messenger, 118,

15Baptiste, P., Le Pape, C., Nuijten, W. 2001

Kluwer Academic Publishers, ISBN 0-7923-7408-8

COIN – COmputational INfrastructure for Opera-tions Research, 2002, http://www.coin-or.org

Cosytec, 2002, http://www.cosytec.comDash Optimization Ltd. 2002: Application of opti-

mization with Xpress-MP, Editions Eyrolles,Paris, France, ISBN 0-9543593-0-8

Diaz, D. 2002, http://gnu-prolog.inria.fr/manual/index.html

Drummond M., Bresina, J., Swanson, K. 1994: Pro-ceedings of the Twelth National Conference onArtificial Intelligence, Seattle, WA, 1098

Giannone, G., Chavan, A. M., Silva, D. R.et al.2000, ASP Conf. Ser., Vol. 216, AstronomicalData Analysis Software and Systems IX, eds. N.Manset, C. Veillet, D. Crabtree (San Francisco:ASP), 111

Grim, R., Jansen, M., Baan, A. et al. 2002, TorbenAnderson, editor, Proceedings of the Workshopon Integrated Modeling of Telescopes, Lund,Sweden, The International Society for OpticalEngineering (SPIE), 51

I2, 2002, http://www.i2.comJohnston, M., Miller, G. 1994, Intelligent Sched-

uling, ed. M. Fox and M. Zweben, San Francis-co: Morgan-Kaufmann, ISBN 1-55860-260-7,391

Silva, D. 2001, The Messenger, 105, 18

View at the ALMA siteon the Zona deChajnantor. The APEXantenna is visible in front of the CerroChajnantor.

Rep

ort

sfr

om

Ob

serv

ers


VISIR, VISIR, AA TTASTEASTE OFOF SSCIENTIFICCIENTIFIC PPOTENTIALOTENTIAL

VISIR, THE ESO-VLT INSTRUMENT MADE FOR OBSERVATIONS IN THE TWO MID-INFRARED ATMOSPHERIC

WINDOWS (THE SO-CALLED N- AND Q-BANDS), IS NOW PRODUCING SCIENTIFIC RESULTS. SOME FIRST RESULTS

ARE DISCUSSED FROM A PEDAGOGIC POINT OF VIEW, EMPHASISING THE VARIOUS MECHANISMS AT WORK THAT

PRODUCE MID-INFRARED RADIATION (THERMAL DUST EMISSION, TRANSIENTLY HEATED DUST EMISSION, ION LINE

EMISSION, PURE ROTATIONAL LINE EMISSION OF MOLECULAR HYDROGEN, SYNCHROTRON EMISSION). THE TWO

KEY ADVANTAGES OF VISIR, I.E., ITS HIGH ANGULAR RESOLUTION AND ITS HIGH SPECTRAL RESOLUTION, ARE

ILLUSTRATED RESPECTIVELY BY THE RESULTS FROM THE OBSERVATIONS OF THE BROWN DWARF BINARY SYSTEM

ε INDI, AND BY KINEMATIC STUDIES OF THE GALAXY NGC 7582.

EERICRIC PPANTINANTIN 1,41,4 , P, P IERREIERRE-O-OLIVIERLIVIER LLAGAGEAGAGE 11, A, ARNAUDRNAUD CCLARETLARET 11, C, CORALIEORALIE DDOUCETOUCET 11,,AANDREASNDREAS KKAUFERAUFER 44, H, HANSANS-U-ULRICHLRICH KKÄUFLÄUFL 44, J, JANAN-W-WILLEMILLEM PPELEL 2,32,3 , R, REYNIEREYNIER FF. P. PELETIERELETIER 22,,

RRALFALF SSIEBENMORGENIEBENMORGEN 44, A, ALAINLAIN SSMETTEMETTE 4,54,5 , M, MICHAELICHAEL SSTERZIKTERZIK 44

1DSM/DAPNIA/SERVICE D’ASTROPHYSIQUE, CEA/SACLAY, SACLAY, FRANCE;2KAPTEYN INSTITUTE, GRONINGEN UNIVERSITY, GRONINGEN, THE NETHERLANDS;

3ASTRON (NETHERLANDS FOUNDATION FOR RESEARCH IN ASTRONOMY), DWINGELOO,THE NETHERLANDS; 4EUROPEAN SOUTHERN OBSERVATORY, 5F.N.R.S. BELGIUM

VVISIR, THE VLT MID-INFRA-red Imager and Spectrom-eter, was installed at UT3(Melipal) in early 2004 andwas successfully commis-

sioned between May and August 2004(Lagage et al., 2004). After this, the time untilits planned start of regular science operationsin period 75 (April–September 2005) hasbeen used to integrate the instrument into theParanal operations and maintenance schemesand to carry out some first Guaranteed TimeObservations (GTO) by the instrument con-sortium and Science Verification (SV) obser-vations to demonstrate the scientific capabil-ities of the instrument. Acomplete list of GTOand SV programs can be found at http://www.eso.org/observing/proposals/gto/visirand http://www.eso.org/science/vltsv/visirsv.The data from the Commissioning and SVobservations are available to the public andcan be retrieved through the ESO science dataarchive (http://archive.eso.org).

Since June 2004, the sensitivities of VISIRwere carefully monitored through a system-atic program of observations of standard starsas often as possible. Figures 1 and 2 show acompilation of these measured sensitivitiesboth in the imaging and the spectroscopicmode. As seen on these figures, the sensitiv-ities are in reasonable agreement with the pre-dicted ones. However, the last 6 months ofmonitoring have shown that the sensitivity ofVISIR depends quite significantly on the con-ditions of the weather and the observationsconditions (seeing, airmass, amount of Pre-cipitable Water Vapour; the value of the lat-ter can be found at http://www.eso.org/gen-fac/pubs/astclim/forecast/meteo/ERASMUS/

par_ fore.txt). Mid-infrared observations withVISIR will thus benefit greatly from the flex-ibility provided by service observing sched-uling since the constraints of the observer willbe much better matched to weather and obser-vation conditions. One should also note thatsome spurious effects, such as detector strip-ing, may degrade the sensitivity performancesporadically.

A sample of first scientific results fromVISIR has been selected to give a first tasteof the overwhelming scientific potential ofthis latest VLT instrument.

VARIOUS ORIGINS OF

MID-INFRARED RADIATION

EMISSION OF LUKEWARM DUST GRAINS IN

THERMAL EQUILIBRIUM

Dust grains (silicates, amorphous carbon,graphite ...) immersed in a radiation fieldreach thermal equilibrium at the temperaturethat corresponds to the balance between ab-sorbed and re-radiated energy. Depending onthe circumstances (orbit, central source lumi-nosity, chemical composition) they can reacha temperature of around 100–500 K. Their re-emitted energy is mostly radiated at infraredwavelengths, such that:

λmax = 2898/T µm, T in K (Wien’s law).

Hence, a warm dust grain at room tempera-ture (i.e., around 300 K), radiates its maxi-mum energy at around 10 µm. Mid-infraredwavelength (10 and 20 µm) windows are thuswell adapted to probe dust grains in orbitsaround a star, typically in the planetary zone(1–50 AU). As a result, the search for foot-

prints left by planets very close to the star industy discs is possible. Figure 3 illustratesthis with images of the dust-disc of β-Pictorisobserved without coronograph (in contrast to visible and near-infrared observations)allowing one to study structures in the inner-most regions; from these observations, themechanism of disc replenishment can beinferred (collisions of planetesimals pro-ducing small particles) (Pantin et al. 1997; Pantin et al. 2005, in preparation) and thestructure of the disc gives precious indica-tions about the presence of gravitational per-turbers, such as massive planets.

Generally speaking, in the mid-IR one isapproaching the Jeans limit. Thus, the con-trast between the stellar photospheres and thecircumstellar environment is more favorableas compared to the near-IR, where corono-graphy is a must.

EMISSION OF VERY SMALL DUST PARTICLES

AND PAHVery small dust grains (e.g. silicates, graph-ite, with sizes smaller than 0.01 µm) as wellas polycyclic aromatic hydrogenated grains(PAH or “big” molecules formed of benzenerings) can be heated quite far away from asource, provided that the source emits a suf-ficient number of visible or ultraviolet pho-tons. These photons stochastically heat thesesmall grains (or big molecules) temporarilyto temperatures close to 1000 K. The grainsthen relax through vibrational modes of C-Hand C-C bonds, at a few precise wavelengthsmainly found in the mid-infrared range (3.3 µm, 6.2 µm, 7.7 µm, 8.6 µm, 11.3 µm ...).

The ISO infrared satellite discovered thatsome isolated Herbig AeBe (HAEBE) stars


Figure1: Imager sensitivity monitored over 6 months, in N-band (left), and Q-band (right). Thesensitivity is estimated for each standard starobservation using the following method. One de-termines the radius at which the signal-to-noise is maximal. The star signal is integrated within thisoptimum radius and the corresponding error iscomputed. The sensitivity is finally deduced fromthe calibrated flux of the star deduced from Cohen et al. (1999) all sky network database ofinfrared calibrators. The horizontal bars representthe theoretical limits of VISIR sensitivity, the black dots show the “best ever” value of sensitivityreached. Note that the two pixel scales available(“small field” and “intermediate field”) have differ-ing sensitivities. Although not expected whenmodeling VISIR sensitivity, “small field” sensitivitiesare systematically better than “intermediate field”ones.

Figure 2: Spectrometer sensitivity monitored over aperiod of 6 months. The left panel shows the low-resolution mode in all four settings currently offeredspanning the full N-band. The right panel displaysthe sensitivity measured in the long-slit [NeII] set-ting at 12.8 µm.

Figure 3: The dust disc of β-Pictoris seen by VISIRat 12.8 and 18.7 µm. The asymmetry (South-Westside brighter than North-East side) is clearly seen.As illustrated here, the longer the observing wave-length, the colder the dust probed. However, thespatial resolution is also worse because of the dif-fraction limit. One key question that will beaddressed with the above images is the following:is there an inner tilt in the β-Pictoris dust disc, asclaimed from Keck observations (Wahhaj et al.2003, Weinberger et al. 2003), or not, as claimedfrom recent Gemini observations (Telesco et al.2005). Some more data processing (image decon-volution) is needed before having the VLT view!

Figure 4: Image of the Herbig Ae star HD 97048.The extension seen in the “PAH” band filter cen-tred on 11.3 µm is clearly visible. Such an exten-sion cannot be explained just by thermal emis-sion of dust grains at an equilibrium temperatureand a population of small grains or PAH mole-cules transiently heated has to be invoked to ex-plain the extension (Doucet et al. 2005). Thespectroscopic observations (see Figure 5) indi-cate that a 11.3 µm feature attributed to PAH is indeed present. As expected the extension at 11.3 µm is larger than that observed recently at 3.3 µm with NACO (Habart et al.,2005, in prepa-ration). A striking feature of the extension is itsasymmetry. Sophisticated models including PAH,flaring disc geometry, viewing angle, inclination …are now needed to interpret these observations.

27© ESO – March 2005Pantin E. et al., VISIR, a Taste of Scientific Potential

Figure 5: Left: Observed spectrum ofHD 97048 around 11.4 µm in thelong-slit low-resolution mode ofVISIR. Right: The measured spatialfull width at half maximum of the HD 97048 spectrum as a function ofwavelength. The PAH emission at11.3 µm is spatially more extendedthan the continuum and confirms the spatial extension of the imageseen in Figure 4.

Figure 6: The Seyfert 2 galaxy NGC 1068. The leftpanel shows the image obtained through the NeIIfilter centered on 12.8 µm, and contains both thecontinuum and the gas emission. The yellow ticksoverplotted mark the local position angle of theisophotes, showing the symmetrical twisting of themid-infrared emission from the center to the out-er parts of the AGN. Also shown is a sketch of thedifferent knots identified in the image. The rightpanel shows the atomic emission line of [NeII] oncethe continuum component has been subtracted.Overplotted scales give the offset coordinates inarcseconds from the central engine position.

Figure 7: The Orion bar observed using VISIR in along-slit, high-resolution setting at 17.03 µm. In the background is shown the 2D spectrum (disper-sion direction is vertical, spatial direction is hori-zontal, North is to the left). The horizontal whitebars overplotted pinpoint the local maxima of theemission line across the field. The yellow curvedisplays the corresponding central wavelengths asa function of the position in the field.

Orion bar (Allers et al. 2004). As shown inthe same figure, the velocity structure of thiscloud might be inferred from the precise cen-tral wavelength of the H2 emission line. Addi-tional checks are needed, however, before de-finitely attributing the observed wavelengthchanges to velocity.

This type of observation from the groundis not easy because of a strong atmosphericemission line very nearby at 17.027 µm; thehigh spectral resolution of VISIR makesVISIR a powerful instrument to detect the H2 line.

Precise wavelength calibration is routine-ly achieved using the sky spectra recorded in the data. The observed atmospheric lines are compared with those predicted from anatmospheric model based on a HITRAN radi-ation transfer model. Using a cross-correla-tion method, an accurate wavelength cali-bration is derived on the fly (see Figure 8).

This observing mode offers great poten-tial for example for the search for cold H2 inprotoplanetary discs.

SYNCHROTRON EMISSION

FROM COMPACT OBJECTS

Charged particles rotating in a strong mag-netic field generate synchrotron emission.The observed spectrum is usually close to

harbour dusty discs in which planetary for-mation is suspected to take place, and showthe signature of such PAH (Waelkens et al.1997, Waters et al. 1998).

One key programme of the VISIR guaran-teed time observations is devoted to the studyof a large sample of such HAEBE pre-main-sequence stars. We have observed one ofthese stars (HD 97048, Figures 4 and 5) usingVISIR in imaging mode (PAH2 filter centredon 11.3 µm) and spectrometry mode (low-res-olution 11.4 µm setting). The 11.3 µm imageshows a quite large extension (2–3 arcsec)that was already suspected from previousobservations (van Boekel et al. 2004). Thespectrum confirms that PAH emission isindeed prominent in this disc, while the rightpanel of Figure 5 proves that the PAH emis-sion is indeed spatially extended.

LINE EMISSION FROM IONIZED GAS

Narrow atomic gas emission lines are alsopowerful probes of the astrophysical condi-tions. The most famous ones in the N-bandare [NeII] at 12.8 µm, [ArIII] at 8.992 µm, and [SIV] at 10.485 µm. Narrow band (R = 50–80) filters corresponding to theselines are available in the VISIR imager andlong-slit mode of the spectrometer. Concern-ing other lines (e.g. [HI] at 12.36 µm, forbid-

den lines such as [NII] at 12.79 µm, or [NaIV]at 9.04 µm), a cross-dispersed mode of thespectrometer will be offered in the future. InFigure 6, we demonstrate the possibility ofstudying the spatial emission of the [NeII] linein the central regions of the Seyfert 2 galaxyNGC 1068. Indeed, after subtraction the 2Dcontinuum emission interpolated fromimages taken through reference filters around12.8 µm, one obtains a map of the pure [NeII]emission (right panel of Figure 6). One cannotice that the [NeII] emission is extended andfollows the Narrow-Line-Regions, but, mostinterestingly, unveils the South-East, dust-extincted component of the ionising cone, as predicted by the unified AGN model (Galliano et al. 2005).

EMISSION LINES FROM PURE ROTATIONAL

MODES OF MOLECULAR GAS

Although this mode has not been offered tothe community yet, VISIR has the capabil-ity to obtain spectra at very high resolution (R ~ 15 000 to 30 000) of “cold” molecularHydrogen H2 at 12.28 and 17.03 µm. The tran-sitions accessible to VISIR correspond to thelowest lying rotational states of the vibra-tional ground-state of Hydrogen. This poten-tial is demonstrated in Figure 7 showing theemission line at 17.03 µm produced in the


Figure 10: ε Indi, the closest binary brown dwarf,observed with VISIR in three filters (PAH1 (8.6 µm),SIV (10.4 µm), and PAH2 (11.3 µm)). We were ableto spatially resolve both components, separated by~ 0.73 arcsec, and determine accurate mid-infra-red photometry for both components independent-ly. In particular, our VISIR observations allowed us to probe the NH3 features in both of the T1(component Ba) and T6 (component Bb) cool browndwarf. For the first time, we could disentangle thecontributions of the two components, and find thatthe cold ε Indi Bb is in reasonable agreement withrecent “cloud-free” atmosphere models. On theother hand, the warmer ε Indi Ba deviates signifi-cantly from any available atmosphere model cal-culations. It may or may not have clouds, and wemight witness non-equilibrium chemical effects of NH3 in ε Indi Bb. One should note that SPITZERcould only measure a composite spectrum (Roelliget al. 2004).

Figure 11: Left: Composite image of Saturn (blue = PAH2 filter at 11.3 µm, red = Q3 filter at19.5 µm). The “shadow” region on the left side of the planet corresponds to ring particles withlower temperatures after cooling down in Saturn’sshadow. Right: A close-up of Saturn’s ringsshowing the D, C, B, and A rings (from left to right),Cassini’s division between the B and A rings, andmore interestingly, an increase of the particle tem-perature (hence increasing thermal emission) in the A ring just before entering the shadow of theplanet.

Figure 8: Wavelengthcalibration of the spec-trometer using a cross-correlation methodbetween the observedsky spectrum (black)and a model of atmos-pheric emission basedon HITRAN computa-tions (blue).

Figure 9: The GalacticCentre observedthrough the PAH1 (8.6 µm) filter. The goalis to detect the syn-chrotron emission fromthe black hole (positionshown by the circle)and constrain its emis-sion models.

through the high spectral-resolution study ofemission of ionized [NeII] gas at 12.8 µm,while the NeII filter image reveals the star-forming, circum-nuclear gas disc in the innerkiloparsec. The precision reached in veloci-ty is around 18 km/s, and the spatial resolu-tion of 0.4 arcsec corresponds to ~ 40 pc.From these data, one can infer an upper limiton the mass of central black hole (Wold et al.2005, in preparation). This example showsthe great potential of VISIR to study gasdynamics at high spatial resolution in the cen-tres of galaxies.

MORE TO COME

VISIR will start science operation in visitorand service mode in P75. Only a selectednumber of instrument modes which could beproperly characterised in the early phases ofcommissioning have been offered for the firstproposal period. Additional modes andimproved sensitivities in imaging and spec-troscopy are planned to be offered for thecoming observing periods. It is worth notinghere that all offered instrument modes arefully supported by the newly developedVISIR (quick-look) pipeline which is avail-able at the telescope and ESO Garching toprocess the data obtained in visitor and ser-vice mode. The latest information about theavailability of the VISIR pipeline can befound at http://www.eso.org/observing/dfo/quality/pipeline-status.html. The latest infor-mation on the status of the instrument isavailable at http://www.eso.org/instruments/visir/.

ACKNOWLEDGEMENTSWe would like to thank the VISIR team at CEA,ASTRON and ESO for all their efforts to build andto bring into operation the VISIR instrument. Weare further very grateful to C. Ferrari, E. Gallianoand M. Wold for making some material availablefor this article prior to publication.

REFERENCESBurrows, A., Sudarsky, D. and Lunine, J. I. 2003,

ApJ, 596, 587Cohen, M. et al. 1999, AJ, 117, 1864Galliano, E., Pantin, E., Alloin, S. and Lagage

P.-O. 2005, MNRAS, submittedGenzel, R. et al. 2003, Nature, 425, 934Lagage, P.-O. et al. 2004, The Messenger, 117, 16Pantin, E. et al. 1997, A&A, 327, 1123Richter, M. J., Jaffe, D. T., Blake, G. A. and Lacy,

J. H. 2002, ApJ, 572, L 161Roellig, T. L. et al. 2004, ApJS, 154, 418Telesco, C. et al. 2005, Nature, 433, 133van Boekel, R. et al. 2004, A&A, 418, 177Waelkens, C., Malfait, K., and Waters, L. B. F. M.,

1997, Ap&SS, 255, 25Waters, L. B. F. M. and Waelkens, C. 1998, ARA&A,

36, 233Wahhaj, Z. et al. 2003, ApJ, 584, L 27


Figure 12: NGC 7582. Left panel:Rotation curve derived from the high-resolution spectrum in the 12.8 µm[NeII] mid-infrared line. Right panel:Image of NGC 7582 obtained in the[NeII] filter showing the slit position(top), and a 2D spectrum unveil-ing the velocity structure (bottom).

Pantin E. et al., VISIR, a Taste of Scientific Potential

photometry of each of the components(SPITZER, although more suited to observesuch faint objects, could only measure thespectrum of the two components and not dis-tinguish each one separately). One can thenput some constraints on the temperature,mass, and radius of each of the components(Sterzik et al. 2005, in preparation).

We observed Saturn’s rings in May 2004,when the opening angle of the rings wasmaximal. VISIR images allow the study ofthe thermal emission from dust particles, andmake it possible to spatially resolve the ringswith a precision never reached before (seeFigure 11) (Ferrari et al. 2005, in preparation).It is amazing that the resolution obtained withVISIR is equivalent to that obtained with thefar-infrared focal plane FP1 of the CIRS(Composite InfraRed Spectrometer) instru-ment on board the Cassini spacecraft, at a dis-tance of 20 Saturn radii during the CASSINI-HUYGENS Tour around Saturn between2004 and 2008. CIRS will observe the ringsat different wavelengths and under differentviewing angles. This makes both instrumentscomplementary. When observing such an ex-tended object it is recommended to observein imaging mode using jitter mode; i.e. someslight offset is applied between two noddingcycles, to avoid a significant number of badpixels (~ 1%) affecting the final image.

HIGH SPECTRAL RESOLUTION: KINEMATICS OF

WARM GAS IN NGC 7582The unique spectral resolution of VISIR in N-(R ~ 30 000) and Q-band (R ~ 15 000) allowsus to resolve kinematically warm or ion-ized gas down to a limit of ~ 15 km/s. One example is shown in Figure 12. Here thedusty, composite starburst-Seyfert 2 galaxyNGC 7582 reveals its kinematic structure

a power-law function (Fν ∝ να with α in the range 1–2) in the radio domain, peakingsometimes in the mid-infrared range as is the case for the black hole in Sgr A (GalacticCentre).

The very high extinction towards this ob-ject (AV close to 30!) makes Sgr A impossi-ble to detect in the visible range and quite dif-ficult in the near-infrared range (Genzel et al.2003).

We have monitored the Galactic Centre(see Figure 9) over four periods in order totry to detect the black hole mid-infrared emis-sion. We could not catch it in its excited state(when it produces flares), but we could derivean upper limit of around 15 mJy for the fluxin its quiescent state (Lagage et al. 2005, inpreparation).

KEY VISIR FEATURES: HIGH ANGULAR

RESOLUTION, HIGH SPECTRAL RESOLUTION

HIGH ANGULAR RESOLUTION: INDIVIDUAL

EMISSION FROM A BINARY BROWN DWARF,SATURN’S RINGS

One key advantage of VISIR over theSPITZER space observatory is the angularresolution. This is illustrated by two exam-ples, the observations of binary brown dwarfsand Saturn’s rings.

Cool stars and L and T brown dwarfs emittheir energy maxima at wavelengths from thenear-infrared up to 20 µm (Burrows et al.,2003), depending on their structure (dust set-tling, presence of “clouds”, etc.) and effec-tive temperature. We had the possibility toobserve a binary brown dwarf during the sci-ence verification phase of VISIR in Novem-ber 2004. As shown in Figure 10, we wereable to detect the binary in the three filterswith which we observed, and estimated the


TTHE VVDS AIMS TO MAP the dis-tribution of galaxies at variousepochs in the life of the uni-verse, in order to trace the evo-lution of galaxies in connection

to the large scale structures. From a well de-fined flux-selected sample of sources iden-tified from deep images, we use VIMOS on the VLT to measure the distance, via the cos-mological redshifts, of a large number of gal-axies and AGNs back in time up to 12 billionyears ago. As we aim to measure the statisti-cal properties of galaxies as free as possibleof biases, we are observing several indepen-

dent fields large enough to be insensitive tolarge fluctuations in the distribution of galax-ies in the universe. The observations of morethan 100 000 galaxies are required to char-acterize the population of galaxies with thebasic statistical measurements like the lumi-nosity function, or correlation function.

Two main surveys make the VVDS, theVVDS-Deep and the VVDS-Wide surveys.The VVDS-Deep is a magnitude limited sur-vey reaching down to IAB = 24 in 1.3 deg2

in two fields, while the VVDS-Wide reachesdown to IAB = 22.5 in 11 deg2 in 4 fields (Table 1). The “Deep” fields are covered with

a 4-passes strategy with VIMOS, leading toa spatial sampling of 1/4th to 1/3rd of the totalI-band limited population observed in pho-tometry, while the “Wide” fields are coveredin a 2-passes strategy. A total number of41000 spectra have been obtained as of Octo-ber 1st, 2004.

We have now completed the First Epochcatalog on the VVDS-02h field and VVDS-CDFS, corresponding to observations ob-tained in the fall of 2002. A total of 11564spectra have been measured. We are in theprocess of completing the redshift measure-ments in the VVDS-Wide fields, for which a

TTHEHE VIMOS-VLVIMOS-VLT DT DEEPEEP SSURURVEYVEY FF IRSTIRST EEPOCHPOCH

OBSEROBSERVVAATIONSTIONS:: EVOLUTIONEVOLUTION OFOF GALAXIESGALAXIES ,,LARGELARGE SCALESCALE STRUCTURESSTRUCTURES ANDAND AGNAGNSS OVEROVER

90% 90% OFOF THETHE CURRENTCURRENT AGEAGE OFOF THETHE UNIVERSEUNIVERSE

THE VIMOS VLT DEEP SURVEY (VVDS) IS A MAJOR REDSHIFT SURVEY OF THE DISTANT UNIVERSE, AIMED AT

STUDYING THE EVOLUTION OF GALAXIES, LARGE SCALE STRUCTURES AND AGNS OVER MORE THAN 90% OF

THE AGE OF THE UNIVERSE. A TOTAL OF 41000 SPECTRA HAVE BEEN OBSERVED SO FAR. FROM THE FIRST

EPOCH OBSERVATIONS CONDUCTED WITH VIMOS, WE HAVE ASSEMBLED ~11000 REDSHIFTS FOR GALAXIES

WITH 0 ≤ Z ≤ 5 SELECTED WITH MAGNITUDE IAB ≤ 24 IN AN AREA 3.1 TIMES THE AREA OF THE FULL MOON.WE PRESENT EVIDENCE FOR A STRONG EVOLUTION OF THE LUMINOSITY OF GALAXIES AND SHOW THAT

GALAXIES ARE ALREADY DISTRIBUTED IN DENSE STRUCTURES AT Z ~ 1.5. THE HIGH REDSHIFT POPULATION OF

~1000 GALAXIES WITH 1.4 ≤ Z ≤ 5 APPEARS TO BE MORE NUMEROUS THAN PREVIOUSLY BELIEVED. AS THE

SURVEY CONTINUES, WE ARE ASSEMBLING MULTI-WAVELENGTH DATA IN COLLABORATION WITH OTHER TEAMS

(GALEX, SPITZER-SWIRE, XMM-LSS, VLA), AS WELL AS EXPANDING TO LARGER SCALES (~100 MPC) TO

PROBE THE UNIVERSE IN AN UNPRECEDENTED WAY.

O. LO. LEE FFÈVREÈVRE 11, G. V, G. VETTOLANIETTOLANI 22, D. B, D. BOTTINIOTTINI 33, B. G, B. GARILLIARILLI 33, V, V. L. LEE BBRUNRUN 11, D. M, D. MACCAGNIACCAGNI 33,,J.-PJ.-P. P. P ICAICATT 44, R. S, R. SCARAMELLACARAMELLA 99, M. S, M. SCODEGGIOCODEGGIO 33, L. T, L. TRESSERESSE 11, A. Z, A. ZANICHELLIANICHELLI 22, C. A, C. ADAMIDAMI 11,,

M. AM. ARNABOLDIRNABOLDI 1100, S. A, S. ARNOUTSRNOUTS 11, S. B, S. BARDELLIARDELLI 55, M. B, M. BOLZONELLAOLZONELLA 5,5,1122, A. C, A. CAPPIAPPI 55, S. C, S. CHARLOTHARLOT 66,,PP. C. CILIEGIILIEGI 22, T, T. C. CONTINIONTINI 44, P, P. F. FRANZETTIRANZETTI 33, S. F, S. FOUCAUDOUCAUD 22, I. G, I. GAAVIGNAUDVIGNAUD 4,4,1111, L. G, L. GUZZOUZZO 77,,

O. IO. I LBERLBERTT 5,5,1122, A. I, A. IOVINOOVINO 77, H.-J. M, H.-J. MCCCCRACKENRACKEN 8,8,1133, B. M, B. MARANOARANO 1122, C. M, C. MARINONIARINONI 77, A. M, A. MAZUREAZURE 11,,B. MB. MENEUXENEUX 11, R. M, R. MERIGHIERIGHI 55, S. P, S. PALALTTANIANI 11, R. P, R. PELLÒELLÒ 44, A. P, A. POLLOOLLO 77, L. P, L. POZZETTIOZZETTI 55,,

M. RM. RADOVICHADOVICH 1100, G. Z, G. ZAMORANIAMORANI 55, E. Z, E. ZUCCAUCCA 55, M. B, M. BONDIONDI 55, A. B, A. BONGIORNOONGIORNO 1122, G. B, G. BUSARELLOUSARELLO 1100,,FF. L. LAMAREILLEAMAREILLE 44, G. M, G. MAATHEZTHEZ 44, Y, Y. M. MELLIERELLIER 8,8,1133, P, P. M. MERLUZZIERLUZZI 1100, V, V. R. R IPEPIIPEPI 1100, D. R, D. R IZZOIZZO 44

1LABORATOIRE D’ASTROPHYSIQUE DE MARSEILLE, FRANCE; 2ISTITUTO DI RADIO ASTRONOMIA, BOLOGNA, ITALY3ISTITUTO DI ASTROFISICA SPAZIALE E FISICA COSMICA, INAF-MILAN, ITALY;

4OBSERVATOIRE MIDI-PYRÉNÉES, TOULOUSE, FRANCE; 5OSSERVATORIO ASTRONOMICO DI BOLOGNA, INAF, ITALY;6MAX PLANCK INSTITUT FÜR ASTRONOMIE, GARCHING, GERMANY; 7OSSERVATORIO ASTRONOMICO DI BRERA, INAF, ITALY

8INSTITUT D’ASTROPHYSIQUE DE PARIS, FRANCE; 9OSSERVATORIO ASTRONOMICO DI ROMA, INAF, ITALY10OSSERVATORIO ASTRONOMICO DI CAPODIMONTE, INAF, NAPLES, ITALY;

11EUROPEAN SOUTHERN OBSERVATORY, GARCHING, GERMANY; 12DEPARTMENT OF ASTRONOMY, UNIVERSITY OF BOLOGNA, ITALY13LERMA, OBSERVATOIRE DE PARIS, FRANCE


total of ~ 27 000 additional spectra have beenobserved.

The VVDS-10h field has now become theHST-COSMOS field with 2 deg2 being ob-served with an unprecedented 590 orbits of HST-ACS. Our primary choice was theVVDS-02h where more than 10 000 spectrawere available at the time of the original HSTproposal, which would have made the red-shifts and morphology of more than 10 000galaxies immediately available. Unfortu-nately, at the pressing request of ESO andHST observatories, the COSMOS field hasbeen set on the VVDS-10h field to avoid asevere time allocation conflict, preventingthe COSMOS-VVDS team to assemble red-shifts in connection with the HST imaginguntil the acceptance of the zCOSMOS pro-gramme, now secured.

The VVDS was defined to make use of the GTO nights allocated to the VIRMOSconsortium in compensation for the substan-tial institute investment in building VIMOS.After the large cut of the GTO from 80 to 50nights, with only 33 nights out of these hav-ing been clear, we are now seeking Large Pro-gram status to complete the VVDS as closeto originally planned.

We present here the VVDS First Epochdata, and the main science results obtained.Results discussed below are computed withH0 = 70, Ωm = 0.3, ΩΛ = 0.7.

DEEP PHOTOMETRY

Excellent deep images are required for a deepredshift survey in order to select targets forspectroscopy in an unbiased way, and to con-strain the spectral energy distribution of theobserved objects. We have completed deepphotometry over the VVDS-02, VVDS-10,VVDS-14 and VVDS-22 fields, in BVR- andI-bands using the CFHT, as well as in U (seeLe Fèvre et al., 2004b, and references there-in) and K (Iovino A., and the VVDS team, inprep.) for smaller areas using ESO tele-scopes. We present an example of our deepimages in Figure 1. The photometric catalogsand images are fully public on http://cencosw.oamp.fr/. We are complete down toIAB = 25, and KAB = 21.7, which ensures thatno bias is propagated from the photometry tothe spectroscopy, for any galaxy type. Re-cently, deep u, g, r, i, and z photometry hasbeen obtained on the VVDS-02h field as partof the CFHT Legacy Survey.

VIMOS MULTI-SLIT SPECTROSCOPY

OBSERVATIONS AND DATA PROCESSING

Multi-object spectra are observed with theLRRED grism with VIMOS on the VLT and1 arcsec slit width. The average slit numberfor the VVDS-Deep and VVDS-Wide is 540and 450 slits, respectively. A pointing patternhas been defined to produce a homogeneousfield coverage for the VVDS-Deep, with agrid of VIMOS pointings offset by (2,2)arcmin, with 4 pointings looking in turn at thesame area in the sky. The VVDS-Wide point-ing strategy allows observing each point inthe sky with two VIMOS pointings produc-ing a roughly homogeneous coverage of thelarge 2 × 2 deg2 fields. In the “Deep” survey,10 exposures of 27 minutes each are taken for a total exposure time of 4.5 h. In the“Wide” mode of the VVDS, 5 exposures of 10 minutes are taken for a total exposuretime of 50 min. The galaxies observed until1 October 2004 are shown in Figure 2.

The data processing has been progressingin two steps: extraction of the wavelength and flux calibrated one dimensional spectra with VIPGI (VIMOS Interactive ProcessingGraphical Interface, Scodeggio et al., 2005),and redshift measurement using KBRED, anautomated tool for redshift measurement(Scaramella et al., in prep). VIPGI integratesin a very efficient environment (Scodeggio et al., 2005) the software delivered to ESO by the VIRMOS consortium for the ESO-VIMOS pipeline. It is available along withcomputing resources at our facilities in Mar-seille and Milan for anybody who obtainsVIMOS data (http://cosmos.mi.iasf.cnr.it/marcos/vipgi/vipgi.html). The spectral reso-lution is R = 230, and the velocity accuracyis 275 km/s as measured from repeated obser-vations of the same objects.

MEASURING REDSHIFTS IN A LARGE RANGE

0 ≤ Z ≤ 5The VVDS is the first survey to assemble acomplete spectroscopic sample of galax-ies based on a simple I-band limit down to IAB = 24, spanning the redshift range 0 ≤ z ≤ 5. Magnitude limited samples have theadvantage of a controlled bias in the selec-tion of the target galaxies, which can lead toa reliable census of the galaxy population asseen at a given rest-frame wavelength.

The redshift domain above redshift 1.3and below the redshift domain of Lyman-break galaxies z > 2.7 has been referred to asthe “redshift desert”. Crossing this “redshiftdesert” is critical to reduce the incomplete-ness of deep redshift surveys, and probe thegalaxy population at an important time in theevolution. As the VVDS observed wave-length range is 5500–9500 Å, it is relativelyeasy to observe spectral features such as[OII]3727 Å up to z ~ 1.4, and then the strongUV features below 1700 Å for z > 2.6, but the redshift range 1.4 < z < 2.7 is more tricky.Using high quality galaxy templates pro-duced from VVDS galaxies to compute red-shifts with KBRED we have been able toidentify a large population of objects with z > 1.4, successfully crossing the “desert” upto z ~ 2.2, as other teams have also recentlybeen doing (e.g. Steidel et al., 2004). Theremaining VVDS “desert” 2.2 ≤ z ≤ 2.6 isdemonstrated to be the result of the selectionfunction imposed by the combination of the faintness of the sources, the wavelengthdomain of the VIMOS-LRRED, and thestrong OH sky emission features. A way toeliminate any possible redshift desert and fol-low the main population of galaxies includ-ing early-type systems would be to combinevisible and near-IR multi-object spectros-

Field

VVDS-0026-04

VVDS-CDFS

VVDS-1003+01

VVDS-1400+05

VVDS-2217+00

Alpha (2000)

02h26m00.0s

03h32m28.0s

10h03m39.0s

14h00m00.0s

22h17m50.4s

Delta (2000)

-04°30(00)

-27°48(30)

+01°54(39)

+05°00(00)

-00°24(27)

Survey area (deg2)

1.3 (Deep)

0.1 (Deep)

4 (Wide)

4 (Wide)

4 (Wide)

Survey mode

Deep

Deep

Wide now HST-COSMOS field

Wide

Wide

Observed as of 1 October 2004

29 deep pointings, ~12 000 slits

5 deep pointings, 1599 redshiftsreleases to the communityhttp://cencosw.oamp.fr

7 pointings, ~2 600 slits



Table 1: VVDS fields

Figure 1: CompositeBVI image from theCFHT12K survey (LeFèvre et al., 2004a).The limiting magni-tude (5σ, in 3 arcsecdiameter aperture) is IAB = 25. This im-age shows only 3 × 4 arcmin2, or1/5000th of the full 16 deg2 imaging survey.


including less reliable flag 1 objects), sam-pling ~ 25 to 30% of the population of galax-ies with IAB ≤ 24. The VVDS data on the CDFS is fully public, with 1599 red-shifts available at http://cencosw.oamp.fr/(Le Fèvre et al., 2004b).

We have assigned to each redshift mea-surement a quality assessment similar to thescheme used by the CFRS (Le Fèvre et al.,1995). Comparing the redshift measurementof 426 objects observed twice and processedindependently, we have been able to esti-mate that the probability of being correct is 53 ± 5 %, 81 ± 3 %, 94 ± 3 %, and 100 %, forgalaxies with flags 1, 2, 3, 4 respectively, andthat the velocity accuracy is 275 km/s. Com-

copy. Unfortunately, the cancellation ofNIRMOS has postponed several critical sci-ence topics in need of statistically significantsamples until KMOS comes into operationafter 2010.

Our observations demonstrate that meas-uring redshifts at z > 1.4 from low resolutionspectroscopy R ~ 200 can be done efficient-ly: a sample spectrum is shown in Figure 3,and high quality VVDS templates are shownin Figure 4 and Figure 10. Low resolutionwith VIMOS allows us to double or triple thenumber of spectra observed in one single ob-servation compared to medium R ~ 600 res-olution and up to 4–5 times more comparedto R ~ 2 500 observations. It is thus particu-larly well suited to assemble very large sam-ples in a reasonable amount of observingtime. Low resolution surveys have the ca-pability to quantify sub-populations, identi-fy rare populations, and establish well de-fined unbiased statistical samples which canthen be followed up at higher spectral re-solution. For programs requiring higher ve-locity accuracy, better sampling and S/N ofnarrow spectral features, medium (~ 600) ormedium high (~ 2 500) spectral resolution canbe very useful indeed, but at a significant lossof total observed spectra.

THE VVDS FIRST EPOCH SAMPLE

The First Epoch sample contains 11 564 ob-jects with spectra; 9 677 are galaxies with aredshift measurement, 836 objects are stars,90 are AGNs, and a redshift could not be mea-sured for 961 objects (Le Fèvre et al., 2005a).There are 1065 objects (galaxies and QSOs)with a measured redshift z ≥ 1.4. When con-sidering only the primary spectroscopic tar-gets, the survey reaches a redshift mea-surement completeness of 78% overall (93%

parison to the dataset of Vanzella et al. (2005)of 39 VVDS galaxies in the CDFS observedwith VLT-FORS2, shows only 9 discrepan-cies, with 8 solved from the better red sensi-tivity of FORS2 and one with a better S/Nfrom the VVDS spectrum. This excellentagreement is statistically fully compatiblewith the distribution of probabilities for theVVDS flags given the small comparison sam-ple from FORS2 and the different observingconditions.

REDSHIFT DISTRIBUTION OF THE

IAB ≤ 24 SAMPLE

The redshift distribution of a spectroscop-ically selected sample of galaxies with

Figure 3: Spectrum ofan absorption linegalaxy with IAB = 23.61,and z = 3.2950, de-monstrating the abilityof low spectral reso-lution (R ~ 230) to mea-sure redshifts fromabsorption line galaxiesdown to the very faintend of the survey andat high redshift.

Figure 4: Example ofgalaxy templatesconstructed from theVVDS, for early type galaxies with 0.6 ≤ z ≤ 1.5.

Figure 2: Field coverage of the VVDS fields as of 1st October 2004 (axis are in degrees): VVDS-02h(top-left), VVDS-10h (top-right), VVDS-14h (bottom-left), VVDS-22h (bottom-right). Galaxies with mea-sured spectra are identified as blue dots, super-imposed on the background of photometric targetswith IAB ≤ 24 for the VVDS-02h field, and IAB ≤ 22.5 for the other fields. The shaded area inthe VVDS-10h represents the COSMOS surveyarea. The VVDS-CDFS is not represented here; itcontains 1599 measured VVDS redshifts (Le Févre et al., 2004b) A total of 14 000 targetshave been observed in the VVDS-Deep, and 27000 in the VVDS-Wide. The VVDS-22h has beencovered once and awaits a second pass of VIMOS observations.

33© ESO – March 2005Le Fèvre O. et al., The VIMOS-VLT Deep Survey First Epoch observations

17.5 ≤ IAB ≤ 24 has been measured for the first time from the VVDS, as shown in Fig-ure 5. The distribution of galaxies peaks at z = 0.8–0.9 (median z = 0.76, mean z = 0.90),while there is a significant high redshift tailextending up to z ~ 5. There are 558 galaxiesin our primary sample with a measured red-shift 1.4 ≤ z < 2.5, 258 with 2.5 ≤ z < 3.5 and161 with 3.5 ≤ z < 5, the largest purely mag-nitude selected sample at these redshifts sofar.

EVOLUTION OF THE LUMINOSITY FUNCTION

AND LUMINOSITY DENSITY FROM Z = 2The evolution of the Luminosity Function(LF) has been investigated from the FirstEpoch VVDS sample, for 17.5 ≤ IAB ≤ 24(Ilbert et al., 2005), using the ALF tool. Weobserve a substantial evolution with redshiftof the global luminosity function in all bandsfrom U to I rest frame, as shown in Figure 6for the B-band. The LFs have been estimatedwithin the absolute magnitude range in whichany kind of galaxy is observable (dashed ver-tical lines in Figure 6, and Figure 7; Ilbert,O., et al., 2004). Compared to the local SDSSvalues, we measure a brightening rangingfrom 1.8–2.4 magnitudes in the U-band, to 0.8–1.4 magnitudes in the I-band, whengoing from z = 0.05 up to z = 2 (Ilbert et al.,2005). The stronger brightening toward bluerrest frame wavelengths suggests that most ofthe evolution of the global LF up to z = 2is related to the star formation history, betterprobed with the luminosity measured at shortrest frame wavelengths.

The First Epoch sample also allows thederivation of the luminosity function for eachof four galaxy types out to z ~ 1.5 (Zucca, E.,and the VVDS team, in prep.). We present inFigure 7 the LF of two extreme galaxy types(among four) that we have used to classifythe galaxies. The difference is striking: galax-ies with early spectral types are only mildlyevolving with a 0.5 magnitudes brightening,while the late-irregular type galaxies were ~ 2 magnitudes brighter and twice more nu-merous at z ~ 1.2–1.5. At bright magnitudesIAB ≤ 22.5, our results are fully consistent withthe CFRS results. Down to IAB ≤ 24, we findsome significant differences compared toprevious surveys based on photometric red-shifts (e.g. Wolf et al., 2003), possibly due tothe degeneracy in photometric redshifts com-putation, demonstrating the need for spectro-scopic redshifts to properly assess the statis-tical properties of the high redshift galaxypopulation.

EVOLUTION OF THE DISTRIBUTION OF GALAX-IES IN LARGE SCALE STRUCTURES FROM Z = 2The First Epoch sample allows a direct esti-mate of the evolution of the spatial distribu-tion of galaxies from within the same survey(Le Fèvre et al., 2005b). The main separationbetween galaxies is best described by thecorrelation length, which we have comput-

ed from the correlation functions ξ(rp,π)and wp(rp) for a total of 7155 galaxies in a 0.61 deg2 area. We find that the correla-tion length r0(z) increases only slightly from z = 0.5 to z = 1.1, with r0(z) = 2.5–2.8 h–1 Mpc(comoving), for galaxies comparable in lu-minosity to the local 2dFGRS and SDSSsamples. This indicates that the amplitude of the correlation function was ~ 2.5 times lower at z ~ 1 than observed at the present time. The correlation length is increasing to

Figure 5: Spectroscopic redshift dis-tribution of a sample of 9141 galaxieswith 17.5 ≤ IAB ≤ 24. This sample is78% complete. The secure redshiftsample is shown as the filled histo-gram, and less secure redshift mea-surements are represented by thedashed histogram. The median red-shift is z = 0.76, while the mean is z = 0.90. The dearth of objects with2.2 < z < 2.6 is a consequence of theobserved wavelength range on theability to measure redshifts, produc-ing the VVDS “redshift desert” (seetext).

Figure 6: Evolution of the luminosityfunction in the rest-frame B-band(Ilbert et al., 2005). Each panel showsthe LF computed in the redshiftrange identified in the lower right cor-ner. Open circles show the VVDSvalues computed using the Vmax andthe continuous line is computedusing the STY technique. The localLF computed by the SDSS is shownas the long dashed line in each panelas a reference.

Figure 7: Evolution of the luminosity function in therest frame B-band for galaxies spectrophotomet-rically classified as early type (left) and late/irregu-lar type (right) (Zucca, E., and the VVDS team, inpreparation). Vertical dashed lines indicate the lu-minosity above which no bias is expected. Whilethe early types show a mild evolution of about 0.5 magnitudes, consistent with passive stellarevolution, the irregular galaxies exhibit a strong 2 magnitudes evolution and an increase in numberdensity of a factor ~2.


r0 ~ 3.0 h–1 Mpc at higher redshifts z = [1.3,2.1], as we are observing increas-ingly brighter galaxies, comparable to the clustering length of galaxies with MB(AB) = –20.5 locally. The slowly varyingclustering of VVDS galaxies as redshiftincreases is markedly different from the pre-dicted evolution of the clustering of darkmatter, indicating that bright galaxies weretracing higher density peaks when the largescale structures were emerging from the darkmatter distribution 9–10 billion years ago, a supporting evidence for a strong evolution of the galaxy vs. dark matter bias. This is thefirst time that a consistent picture of galaxyclustering is obtained over such a large timebase from within the same survey as shownin Figure 8. The analysis of the clustering of sub-populations (e.g. selected by type or luminosity) is in progress (Meneux et al.,Guzzo et al., Pollo et al., in preparation).

The complete spectroscopic survey sam-ples the high redshift universe at z ~ 1 with amean inter-galaxy separation of ~ 5 Mpcequivalent to the sampling of the 2dF galaxyredshift survey at a mean redshift z ~ 0.1. Thegalaxy density field can be computed reliablyup to z ~ 1.5. It spectacularly shows strongover-densities (e. g. shown in Figure 9) alter-nating with low density regions, very densewalls, and filamentary distribution, whichcan be traced in the entire survey cone, andover transverse scales ~ 30 Mpc. The size ofstructures observed indicates that probingeven larger scales ~ 100 Mpc is necessary, asplanned in the original VVDS.

The second moment of the ProbabilityDistribution Function (PDF) of galaxy over-densities stays roughly constant up to z ~ 1.5,with σ8 = 0.94 ± 0.07 (in redshift space). Thethird moment of the PDF, the skewness,increases with cosmic time: we find that theprobability of having under-dense regions ishigher at z ~ 0.7 than it was at z ~ 1.5. By com-paring the observed PDF of galaxy fluctua-tions with the theoretically predicted PDF ofmass fluctuations, we find that the galacticbias is increasing by ~ 40% from z ~ 0.7 to z = 1.5. Red objects appear more clusteredthan blue objects at all epochs investigated.The relative bias between the density fluctu-ations of red and blue objects is constant as afunction of redshift with a relative bias of ~ 1.4 comparable to the local z = 0 value.These results are presented in Marinoni et al.(submitted).

THE HIGH REDSHIFT

1.4 ≤ Z ≤ 5 VVDS POPULATION

There are more than 970 galaxies with red-shifts measured in the range z = [1.4,5] in theprimary sample of the First Epoch VVDS.The composite spectra of galaxies with mea-sured redshifts in the range z = [1.4,2.5], z = [2.5,3.5], and z = [3.5,5], are shown in Figure 10. These galaxies are bright, withabsolute magnitudes MB(AB) = –21.5 or

Figure 8: Evolution of the correlationlength r0(z) along cosmic time (Wolf et al., 2003). The filled circles are theVVDS measurements in the VVDS-02h field, while the open squares aremeasured in the VVDS-CDFS field.The VVDS-CDFS point at z ~ 0.7shows a higher correlation lengthbecause of the presence of a strongover-density at this redshift.

Figure 9: Galaxy density field arounda large “wall” of galaxies in theVVDS-02h field. This dense structureis at a redshift z = 0.97, it is rela-tively thin and extends over the fulltransverse cone sampled so far or ~ 40 Mpc. Extending the survey up to2 degrees size would allow us toprobe such large scale structures onscales ~ 100 Mpc.

Figure 10: Composite VIMOS spectraof the high redshift VVDS First Epoch galaxy sample in each of theredshift domains z = [1.4,2.5], z = [2.5,3.5], z = [3.5,5]. UV featuresare clearly identified, and the blue continuum indicates activelystar forming galaxies.

brighter. They show a strong UV continuumindicating vigorous star formation (Figure10). From the rest frame 1500 Å UV flux wecan infer that these galaxies produce starswith a star formation rate of several tens ofsolar masses per year (uncorrected for extinc-tion). The volume density of galaxies at theseredshifts seems to be higher than previouslyreported for other galaxy populations at thesame redshifts. This is under investigationand will be reported elsewhere (Le Fèvre, O.,and the VVDS team, in prep.), together withthe detailed properties of these galaxies (Paltani, S., and the VVDS team, in prep.).

THE FAINT AGN POPULATION OUT TO Z ~ 5Since the only selection criterion for theVVDS is magnitude, with no additional cri-

teria based on morphology or colors, theVVDS is also an ideal survey for selecting acomplete and unbiased sample of AGN. Wehave identified 133 broad line QSOs up tonow, 76 identified in the VVDS-Deep sur-vey of the VVDS-02h and VVDS-CDFS, and 57 QSOs identified so far in the VVDS-Widesurvey of the VVDS-10h and VVDS-22h.The redshift distribution is shown in Fig-ure 11. The VVDS probes ~ 2–3 magnitudesfainter than the 2dF QSO survey, extendingthe measurement of the LF toward the faintend, and sampling a transition regime be-tween bright AGNs and galaxies, where stan-dard selection for AGN samples is sig-nificantly incomplete. This is presented inGavignaud et al. (in prep.), and Bongiorno etal. (in prep).

35© ESO – March 2005Le Fèvre O. et al., The VIMOS-VLT Deep Survey First Epoch observations

MULTI-WAVELENGTH OBSERVATIONS

We have been conducting observations atwavelengths outside the optical and near-infrared window of the VVDS, and we havestimulated observations from other teams.We have observed the VVDS-02h field withthe VLA at a depth 17 µJy (1σ) (Bondi et al.,2003, and Ciliegi, P., et al., in prep.). TheVVDS-02h field is observed in the UV withGALEX, in the mid-IR with Spitzer by theSWIRE team, in the X-rays with XMM by theXMM-LSS team. The VVDS-22h field isobserved by GALEX, and the VVDS-10h,now the COSMOS field, is observed at allwavelength from the radio to the X-ray band(http://www.astro.caltech.edu/~cosmos/).The immediate availability of ~1000 VVDSredshifts of UV sources detected withGALEX in the VVDS-02h has allowed pub-lishing for the first time the luminosity func-tion, luminosity density, and star formationrate directly measured at 1500 Å rest frame(Arnouts et al., 2004).

The SWIRE team has observed theVVDS-02h field with Spitzer, at wavelengths3.6, 4.5, 5.8, 8 and 24 microns. The joint cat-alogue of SWIRE and optical data gives red-shifts for more than 3200 sources at 3.6 µm,and more than 220 sources at 24 µm. Thecombination of deep optical, near-IR andmid-IR data will be very powerful to probethe evolution of stellar light, and dustenshrouded star formation.

A detailed dynamical study of VVDSgalaxies with 1 < z < 2 will be conducted with

Figure 11: (left) Redshift distribution of broad lineAGNs identified in the VVDS-Deep IAB ≤ 24 sample(top), and in the VVDS-Wide IAB ≤ 22.5 sample(bottom). QSOs with several features supportingthe redshift are represented by the shaded his-togram, while the light grey histogram indicatesQSOs with only one secure broad line observed.(right) Example of QSO spectra covering our red-shift and magnitude range.

3D spectroscopy and SINFONI (Lemoine-Buserolle et al.). The VVDS is a uniquelysuited sample to perform an unbiased sampleselection for follow-up spectroscopy, in par-ticular to measure star formation indicatorslike the Hα line in emission, from near-IRspectroscopy. A UK team led by A. Bunker isproposing to use CIRPASS as a visitinginstrument on the VLT to perform near-IRmulti-fiber spectroscopy.

PROSPECTS

The VVDS is now bringing into full view thedistant universe of normal galaxies like ourMilky Way and its progenitors over a largetime of evolution. The hard but rewardingapproach of a magnitude limited sampleallows a statistically robust measurement ofthe evolution of the galaxy population usingthe luminosity function, correlation function,or the probability distribution function ofdensity fluctuations. A wealth of new resultswill appear in the months to come from theFirst Epoch catalogue of ~ 10 000 redshiftsand the ~ 30 000 spectra being processed. Itis essential to our understanding of galaxyevolution that the VVDS be completed asoriginally planned with (i) an increase of thearea covered in the VVDS-02h “deep” fieldby at least a factor 2, and (ii) a complete red-shift measurement in ~ 10 deg2 in the fullVVDS-14h and VVDS-22h. This will extendthe survey to ~ 100 Mpc at z ~ 1–5, unprece-dented up to now, probing scales larger thanthe mean size of large voids and dense struc-

tures. This will undoubtedly bring new in-sight into the complex interplay between gal-axy evolution and the underlying large scalestructure distribution. The VVDS, combinedwith the zCOSMOS survey connecting theevolution of galaxy morphology to the largescale structures, is establishing Europeanastronomy as a leading authority in high sta-tistical accuracy deep redshift surveys, a cen-tral key to our understanding of galaxy evo-lution.

REFERENCESArnouts, S. et al. 2004, ApJ, 619, 43Bondi, M. et al. 2003, A&A, 403, 857Ilbert, O. and the VVDS team 2004, MNRAS, 351,

541Ilbert, O., Tresse, L., Zucca, E. and the VVDS team

2005, A&A, in press, astro-ph/0409134Le Fèvre, O., Crampton, D., Lilly, S. J. et al. 1995,

ApJ, 455, 60Le Fèvre, O., Mellier, Y., McCracken, H. J. et al.

2004a, A&A, 417, 839Le Fèvre, O., Vettolani, G. and the VVDS team

2004b, A&A, 428, 1043Le Fèvre, O., Vettolani, G., Garilli, B., Tresse, L.

and the VVDS team 2005a, A&A, in press, astro-ph/0409133

Le Fèvre, O., Guzzo, G., Meneux, B. et al. 2005b,A&A, in press, astro-ph/0409135

Scodeggio, et al. 2005, PASP, submitted, astro-ph/0409248

Steidel, C. C., Shapley,A. E., Pettini, M. et al. 2004,ApJ, 604, 534

Vanzella, E., Cristiani, S. 2005, A&A, in press,astro-ph/0406591

Wolf, C. et al. 2003, A&A, 401, 73


RRECENTECENT AASTROPHYSICALSTROPHYSICAL RRESULESULTSTS

FROMFROM THETHE VLVLTITI

WE PRESENT A REVIEW OF RECENT ASTROPHYSICAL RESULTS BASED ON DATA OBTAINED WITH THE ESO VERY

LARGE TELESCOPE INTERFEROMETER (VLTI). THE VERY FIRST VLTI RESULTS HAVE ALREADY BEEN REVIEWED

BY RICHICHI & PARESCE (2003). REMARKABLY, THE FIRST SCIENTIFIC RESULTS FROM THE MID-INFRARED

INSTRUMENT MIDI, WHICH EMERGE FROM SCIENCE DEMONSTRATION AND GUARANTEED TIME OBSERVATIONS,ALREADY COVER A VERY BROAD RANGE OF ASTROPHYSICAL TOPICS. THESE TOPICS INCLUDE ACTIVE GALACTIC

NUCLEI, YOUNG STELLAR OBJECTS AND PROTOPLANETARY DISCS, THE ENVIRONMENTS OF HOT STARS, AS WELL

AS EVOLVED STARS AND THEIR ENVELOPES.

MMARKUSARKUS WWITTKOWSKIITTKOWSKI 11, F, FRANCESCORANCESCO PPARESCEARESCE 11, O, OLIVIERLIVIER CCHESNEAUHESNEAU 22, P, P IERREIERRE KKERERVELLAVELLA 33,,AANTHONYNTHONY MMEILLANDEILLAND 22, K, KLAUSLAUS MMEISENHEIMEREISENHEIMER 44 ANDAND KKEIICHIEIICHI OOHNAKAHNAKA 55

1EUROPEAN SOUTHERN OBSERVATORY, GARCHING, GERMANY2OBSERVATOIRE DE LA CÔTE D’AZUR, GRASSE, FRANCE

3LESIA, OBSERVATOIRE DE PARIS, FRANCE4MAX-PLANCK-INSTITUT FÜR ASTRONOMIE, HEIDELBERG, GERMANY5MAX-PLANCK-INSTITUT FÜR RADIOASTRONOMIE, BONN, GERMANY

OOBSERVATIONS WITH THE ESOVLT Interferometer startedwith the achievement of First Fringes in March 2001,which employed the test

siderostats and the commissioning instru-ment VINCI. Since then, the VLTI instrumen-tation has been constantly expanded, and sci-entific observations have continuously beencarried out using the instrumentation that hadbeen available at any given time. A report onthe technical status of the VLT Interferome-ter was for instance presented by Glindemannet al. (2004). A huge number of scientifical-ly interesting VINCI data were secured be-tween March 2001 and September 2004 with-in commissioning and early shared riskprogrammes, and all of these VINCI data arepublicly available1. Early MIDI observationsin the framework of science demonstrationand guaranteed time programmes have beenconducted since 2003. The data based onMIDI science demonstration programmes arealso publicly available2. The MIDI instru-ment has been offered to the whole astro-nomical community for general programmessince ESO period P73 (observations startingin April 2004). First science demonstra-tion and guaranteed time observations using the near-infrared phase closure instrumentAMBER were conducted in October andDecember 2004.

The first refereed publications based on dataobtained with the VLTI appeared in the year2003. A review of the very first VLTI resultswas presented by Richichi & Paresce (De-cember 2003). Since then many new resultshave been published that are based on data obtained with VLTI instruments. TheESO telescope bibliography lists for the years2003 and 2004 a total of 21 refereed publica-tions directly using ESO data, which wereobtained with the VLTI instruments VINCIand MIDI (see Table 1). These data weretaken within VINCI commissioning, VINCIshared risk, MIDI science demonstration, andMIDI guaranteed time programmes. Thesepublications are listed in the references be-low. Also listed are 11 additional publicationsthat appeared in 2005 or that are in press, andthat qualify to be included in the ESO tele-scope bibliography. Here, we review thesescientific results, concentrating on those thatwere published after the review of first VLTIresults by Richichi & Paresce (2003). We puta particular emphasis on the results emerg-ing from MIDI science demonstration andguaranteed time programmes (Jaffe et al.2004; Leinert et al. 2004; van Boekel et al.2004; Chesneau et al. 2005 a, b, c; Ohnaka etal. 2005). Recently, Richichi et al. (2005) pre-sented an updated version of CHARM (cata-logue for high angular resolution measure-ments), a catalogue that lists high angularresolution measurements in general, includ-ing results from the VLTI.

The VLTI results reviewed in this arti-cle illustrate that the VLT Interferometerprovides excellent opportunities to address a

broad range of important astrophysical top-ics. With the large collecting areas of the 8 m VLT Unit Telescopes, VLT Interferome-try is not limited to the brightest stars, as mostinterferometric facilities so far, but can beused to study relatively faint sources. Resultsobtained so far include already, for instance,studies of the innermost cores of Active Ga-lactic Nuclei (AGN) harbouring supermas-sive black holes, the origins of stars such asour sun and the formation of protoplanetarydiscs, fundamental parameters of stars in evo-lutionary stages such as our sun, the evolu-tion and fate of stars and their mass-loss tothe circumstellar environment, the environ-ment of hot stars, and distance estimatesbased on measurements of Cepheid pulsa-tions.

Remarkably, several of the publicationsreviewed here include results that were ob-tained by combing observations using differ-ent instruments and facilities. For instance,Chesneau et al. (2005a) combined obser-vations obtained with MIDI, optical spec-tra taken with the HEROS instrument, andinfrared spectra from the 1.6 m Braziliantelescope. Kervella et al. (2003b), Pijpers et al. (2003), and di Folco (2004) combined VINCI data with asteroseismic observations.Boboltz & Wittkowski (2005) conductedcoordinated near-infrared (VINCI) and radio(VLBA) interferometry. Woodruff et al.(2004) combined VINCI observations with near-infrared speckle observations.Wittkowski et al. (2004) combined VINCIinterferometry with available spectropho-tometry. These examples show that interfer-

1 www.eso.org/projects/vlti/instru/vinci/vinci_data_sets.html

2 www.eso.org/projects/vlti/instru/midi/midi_data_sets.html


Year

2003

2004

2005 (so far)

VINCI

6

12

8

26

MIDI

–

3

4

7

Total

6

15

11

32

Table 1: Refereed pub-lications based on data obtained with VLTIinstruments. Source:ESO Telescope Bibliog-raphy.

Figure 1: Top panel: Model dust distribution in thenucleus of NGC 1068 as inferred from the MIDIobservations (North is at top, East to the right). Acompact hot component (yellow) is embedded into the well resolved warm component (red).Hatching indicates the depth of the SiO absorp-tion. Bottom panel: The VLBA radio continuummap at 5 GHz of the S1 component in NGC 1068(from Gallimore et al 2004.) superimposed on themodel. Note the widening of the radio componentat its western edge.

Figure 2: First K-band long-baseline interferometryof NGC 1068, obtained with VINCI. The VINCIvisibility point at a baseline of 46 m indicates thatpart of the K-band flux originates from very small scales of less than 5 mas or 0.4 pc. Two pos-sible two-component model visibility curves areshown which are consistent with this VINCI visibil-ity measurement and with speckle observations up to a baseline of 6 m. These models consist of a0.1 mas and 3 mas compact core, respectively,plus a more extended component. Based onWittkowski et al. (2004b).

ometry, and in particular VLT Interferometry,is becoming a very useful complementarytool for general astrophysics among the manytools that are available to the astronomicalcommunity. It can be expected that more syn-ergies between observations with the VLTIand external instruments and facilities willemerge, ultimately also including ALMAandOWL.

GALACTIC NUCLEI

VLTI observations of Active Galactic Nuclei(AGN) with MIDI started soon after the in-strument was ready for scientific observa-tions with two UTs: the first interferomet-ric spectrum of the famous Seyfert II gal-axy NGC 1068 was obtained during sciencedemonstration time in June 2003. A secondset of MIDI measurements were carried outin November 2003. Also in November 2003,the first near infrared K-band long-baselineinterferometric observations of NGC 1068succeeded, which employed the VINCI in-strument after adaptive optics wavefront cor-rection (MACAO). The first successful MIDIobservation of the second brightest galacticnucleus, that of the Circinus galaxy, becamepossible in February 2004. Two more mea-surements with baseline UT2–UT3 were obtained during guaranteed time in June.

The Circinus nucleus is over three timescloser and ten times less luminous than NGC1068. The analysis of the data shows that thecentral dust structure is well resolved at 10µm even with a 43 m baseline. Nevertheless, due to the small distance, its physical size isonly 1 pc diameter. The observed visibilitiesas a function of projected baseline orientationhint at a much flatter dust distribution thanthat observed in NGC 1068 (see below).

For the purpose of this overview, however,we focus on the discussion of the VLTI resultsof NGC 1068 since they best demonstrate thewealth of information which already has beenprovided by the VLTI. A first account of theMIDI observations of NGC 1068 was pub-lished by Jaffe et al. (2004). On the basis ofthe interferometric spectra covering the range8 µm < λ < 13 µm taken with 78 m and 42 mbaseline, respectively, the authors argue thattwo central components are discernible: awell resolved, warm (T = 320 K) compo-nent of 3 pc diameter which is geometrically thick (h/r Q 0.6), and a very compact hot (T > 800 K) component, which is marginallyresolved along the source axis. The SiOabsorption feature increases in depth whenusing higher spatial resolution, i.e. longerbaseline, suggesting that the central hot com-ponent suffers strong dust extinction equiva-

lent to AV > 50. The model is visualized inFigure 1 (top).

A natural interpretation of these findingsidentifies the 320 K component with theAGN-heated dusty torus expected in the coresof Seyfert galaxies, while the embedded hotcomponent represents the wall of the funnelwhich is heated almost to the dust subli-mation temperature and confines the well studied outflows along the source axis of NGC 1068. Additional support for this pic-ture comes from radio continuum observa-tions with the VLBA presented by Gallimoreet al. (2004): Superimposing their map of the central component S1 on the MIDI model(Figure 1 bottom) shows that the disc-likeradio source coincides in size with the hotcomponent. Presumably, in the radio contin-uum we observe free-free emission of the ion-ized gas within the funnel which is continu-ously released from the walls and eventuallyfeeds the central accretion disc. It is intrigu-ing to identify the increased height of theradio continuum source at its outer parts withthe ionized gas which has just been “eatenaway” from the funnel walls by the intenseUV radiation from the central engine.

At the same time,Wittkowski et al. (2004)presented the first K-band long-baseline in-terferometric measurement of the nucleus ofNGC 1068 (see above). A squared visibilityamplitude of 16.3 ± 4.3 % was measured forNGC 1068 at a baseline length of 46 m. Thisvalue corresponds to a FWHM of the K-bandintensity distribution of 5.0 ±0.5 mas if it con-sists of a single Gaussian component. Takinginto account K-band speckle interferom-etry observations up to a baseline of 6 m (Wittkowski et al. 1998), a multi-componentmodel is suggested for the intensity distribu-tion where a part of the flux originates fromscales clearly smaller than 5 mas (< 0.4 pc),and another part of the flux from larger scales.Figure 2 shows the measured VLTI/VINCI K-band visibility together with possiblevisibility models. For illustration, two dif-ferent two-component visibility curves are shown which are consistent with both, theVLTI/VINCI visibility measurement at abaseline of 46 m, and the speckle observa-tions up to a baseline of 6 m. These modelsconsist of a 0.1 mas or 3 mas compact core,respectively, plus a more extended compo-nent.

Taking the MIDI and VINCI results ofNGC 1068 together, there remains a com-pletely open question: How are the compactcomponents at 2 and 10 µm related to eachother? Without an accurate (relative) astrom-etry (to about 1 mas) available yet, we are left

to speculations based on indirect arguments.From the depth of the SiO feature one infersa K-Band extinction toward the hot MIDIcomponent of AK > 12. One possible expla-nation is a very porous dust torus which occa-sionally leaves holes of low extinction towardits hot centre. The compact K-band compo-nent could then represent a direct view to thecentral source or the inner part of the funnelviewed through such a hole. Alternatively, theK-band component could be located some-what above the torus plane: in this case it isvery unlikely that dust can be heated to > 1200 K (unless the UV radiation from thecentral source is extremely collimated). Moreprobably, the K-band light is then scatteredby dust clouds above the torus.


Pott et al. (this issue, page 43) describe indetail their program to study the stellar pop-ulation close to our Galactic Centre. In par-ticular, they describe the first successfulMIDI observations of the dust enshroudedstar IRS 3 located in the central lightyear ofour Galaxy. Such studies aim at investigat-ing the star formation history and the inter-action of stars with their environment in thepresence of tidal forces exerted by the grav-itational potential of the central supermassiveblack hole.

In summary, it is obvious that already thefirst one and a half years of VLTI observa-tions of galactic nuclei have provided com-pletely new views into the cores of these stillmysterious objects. New facilities like thephase-closure instrument AMBER or the ex-ternal fringe tracking system FINITO willsoon extend the VLTI capabilities to a levelthat interferometric observations of galacticnuclei will become both routine and the mostimportant driver for our understanding ofAGN physics.

SIZES AND MINERALOGY OF CIRCUMSTELLAR

DISCS AROUND YOUNG HERBIG AE/BE STARS

MIDI on the UTs was recently used to de-termine the characteristic sizes of HerbigAe/Be star discs (Leinert et al. 2004) and theirdetailed dust properties (Van Boekel et al.2004). The characteristic 10 µm sizes of thediscs around seven Herbig Ae/Be stars turnedout to be in the range of 1–10 AU and cor-related with the slope of their SED between 10 and 25 µm (Figure 3). This correlationtends to give observational support even forsuch a limited sample to the hypothesis thatHerbig Ae/Be star discs can be distinguishedby whether they are flaring or non-flaring. Inthis context, one would expect, that as ob-served, the reddest objects should be largersince the flaring disc geometry exposes moredistant material to direct illumination than thenon-flaring geometry.

The detailed spectral shapes of three starsof this sample (HD 142527, HD 144432, andHD 163296) were then studied as a functionof spatial scale in order to better understandthe properties of circumstellar dust in thesevery young stars. It has been known for sometime that most of the dust in discs around new-born stars is made up of silicates. In the natalcloud this dust is amorphous, i.e. the atomsand molecules that make up a dust grain areput together in a chaotic way, and the grainsare fluffy and very small, typically about 10–4 mm in size. However, near the young starwhere the temperature and density are high-est, the dust particles in the circumstellar disctend to stick together so that the grains be-come larger. Moreover, the dust is heated bystellar radiation and this causes the moleculesin the grains to re-arrange themselves in geo-metric (crystalline) patterns.

Accordingly, the expectation is that thedust in the disc regions that are closest to

Figure 3: Correlation between the mid-IR spectralslope and the half light radius corresponding to the observed visibilities at 12.5 µm. The largestsources are those with the reddest mid IR SED.From Leinert et al. (2004).

the star is soon transformed from ”pristine”(small and amorphous) to ”processed” (larg-er and crystalline) grains. Model calculationsshow that crystalline grains should be abun-dant in the inner part of the disc at the timeof formation of the Earth. In fact, the mete-orites in our own solar system are mainlycomposed of this kind of silicate. Spectralobservations of silicate grains in the mid-infrared around 10 µm should tell whetherthey are ”pristine” or ”processed”. Earlier ob-servations of discs around young stars haveshown a mixture of pristine and processedmaterial to be present, but it was impossibleto tell where the different grains resided inthe disc.

Thanks to a hundred-fold increase in angu-lar resolution with the VLTI and the highlysensitive MIDI instrument, detailed infraredspectra of the various regions of the pro-toplanetary discs around the three youngHerbig Ae stars, only a few million years old,now show that the dust close to the star ismuch more processed than the dust in theouter disc regions. In one star (HD 142527)the dust is processed in the entire disc (Fig-ure 4). In the central region of this disc, it isextremely processed, consistent with com-pletely crystalline dust. In the other two stars(HD 144432 and HD 163296) the dust in theinner disc is fairly processed whereas the dustin the outer disc is nearly pristine (Figure 5).

An important conclusion from the VLTIobservations is, therefore, that the buildingblocks for Earth-like planets are present incircumstellar discs from the very start. Thisis of great importance as it indicates that plan-ets of the terrestrial type like the Earth aremost probably quite common in planetarysystems, also outside the solar system.

The present observations also have impli-cations for the study of comets. Some – per-haps all – comets in the solar system do con-tain both amorphous and crystalline dust.Comets were definitely formed at large dis-tances from the Sun, in the outer regions ofthe solar system where it has always beenvery cold. It is therefore not clear howprocessed dust grains may end up in comets.In one theory, processed dust is transportedoutward from the young Sun by turbulence inthe rather dense circumsolar disc. Other the-ories claim that the processed dust in cometswas produced locally in the cold regions overa much longer time, perhaps by shock wavesor lightning bolts in the disc, or by frequentcollisions between bigger fragments.

The MIDI observations now imply that thefirst theory is the most likely explanation forthe presence of processed dust in comets. Thisalso implies that the long-period comets thatsometimes visit us from the outer reaches ofour solar system are truly pristine bodies, dat-ing back to an era when the Earth and the otherplanets had not yet been formed. Studies ofsuch comets, especially when performed in-situ, will therefore provide direct access to

the original material from which the solar sys-tem was formed.

HOT STARS AND THEIR ENVIRONMENT

Be stars are hot stars that exhibit Balmer linesin emission and free-free infrared excess, in-terpreted as due to a compact equatorial gas-eous disc around these objects. Be stars arerelatively frequent among the B-type objects.α Ara (HD 158427, B3Ve), which is one ofthe closest Be stars with an estimated distanceof 74 pc (Hipparcos parallax) and an infraredexcess among the highest of its class, has beenchosen for the first MIDI observations of a Be star (Chesneau et al. 2005a). Surpris-ingly, α Ara could not be resolved with the 102 m projected baseline of the UT1 andUT3 telescopes, putting an upper limit to

Figure 4: The mid-IR spectrum of the inner regionof the protoplanetary disc around the young starHD 142527, as observed with MIDI (upper panel).Below it are shown laboratory spectra of two crys-talline minerals as well as of an Interplanetary DustParticle (IDP; captured in the Earth’s upper atmos-phere) with hydrated silicates and, at the bottom, a typical telescopic spectrum of dust grains in theinterstellar space. The spectral “signatures” ofcrystalline pyroxene and olivine, i.e. peaks at wave-length 9.2 and 11.3 µm, respectively, are clearlyvisible in the spectrum of the inner stellar disc, de-monstrating the presence of these species in thatregion of the disc. From Van Boekel et al. (2004).


Figure 6: Schematicview of the proposedmodel for the Be star α Ara and its circum-stellar environment.From Chesneau et al.(2005).

Figure 7: Schematicview of a Mira variablestar.

Figure 5: Schematic view of a circumstellar discand the observed MIDI-spectra of the inner and outer regions of the discs around three youngHerbig Ae stars, HD 163296, HD 144432 and HD 142527 (black lines). In all of them, there areclear spectral differences between the inner and outer regions, indicating a difference in miner-alogy. The general broadening of the spectral“mountain” in the inner discs is a sign of largergrains and the spectral peak at wavelength 11.3 µm indicates the presence of crystalline sili-cates (cf Figure 4). Also shown are best-fit model spectra (red lines), based on mixtures of thementioned mineral species. From Van Boekel et al. (2004).

the disc size in the Ν-band of the order of φmax = 4 mas ± 1.5, i.e. 14 RV at 74 pc andassuming RV = 4.8RA, based on the spectraltype.

On the other hand, a high density of thedisc is mandatory in order to reproduce thestrong Balmer emission lines and the infraredexcess. Optical spectra from the HEROS in-strument, and infrared ones from the 1.6 mBrazilian telescope have been used togetherwith the MIDI visibilities to constrain the sys-tem parameters. Using the circumstellar pa-rameters from the SIMECA code, a goodagreement was found with spectroscopic andinterferometric data only if the disc is some-how truncated at about 25 RV. The authorsfound variations of the hydrogen recombina-tion line profiles and of the radial velocitiesof α Ara, and they argue that this may be anindication for a possible low-mass compan-ion that could cause the truncation of the disc.

Using the stellar parameters from the mod-el, i.e. a mass of 10 MA, a 70 day period wouldgive a radius of about 32 stellar radii, a valuein agreement with the estimate based on theMIDI data for a disc truncated at 25 RV, i.e.somewhat smaller than the companion orbit.Figure 6 shows a schematic view of the pro-posed model for the Be star α Ara and its cir-cumstellar environment, including the possi-ble unseen companion.

Previously, Domiciano de Souza et al.(2003) reported on VINCI observations of the

rapidly rotating Be star Achernar. They foundan oblateness of Achernar which exceeds the-oretical predictions and, hence, puts new per-spectives on our understanding of the Be phe-nomenon. α Ara presents a rotational veloc-ity v sin i which is even larger than that ofAchernar, and is, hence, likely to be very dis-torted, as well. The oblateness of the surfaceof α Ara can be studied with AMBER, andthis will likely provide additional implica-tions on the wind model.

The famous Luminous Blue Variable(LBV) η Car was recently studied byChesneau et al. (2005b) by combining MIDIobservations and adaptive optics measure-ments using NACO at the VLT. They reacheda spatial resolution of 60–80 mas with decon-volved NACO images in the L-band. MIDI ininterferometric mode provided a mean reso-lution of about 20 mas. Using this combina-tion of MIDI and NACO observations, theauthors were able to study the clumpiness ofthe nebula as well as to isolate the flux of thecentral star. The latter enabled them to con-strain the spectral energy distribution (SED)of the central source by itself for the wave-length range from 2.2–13.5 µm.

STARS ON THE

ASYMPTOTIC GIANT BRANCH (AGB)Interferometry is known to be a valuable toolfor studying fundamental stellar parameters,the atmospheric structure, and the circumstel-lar environment of AGB and post-AGB stars.The spectro-interferometric capabilities of

the VLTI instruments at near (AMBER) andmid-infrared (MIDI) wavelengths are expect-ed to lead to new important insights into thisstage of stellar evolution, and in particularinto the atmospheric structure, the mass lossprocess, the formation of circumstellar enve-lopes, and the evolution of AGB stars towardplanetary nebulae. Figure 7 shows a schemat-ic diagram of a typical Mira variable and itssurrounding environment. The various re-gions shown in this diagram can be probed bycombining observations from various inter-ferometric instruments and facilities in a mul-tiwavelength study. In particular, mid-infra-red interferometric observations are adequatefor probing the outer atmosphere, i.e. the re-gion between the top of the photosphere andthe dust envelope where mass outflows areexpected to be initiated, and the expandingdust shell.

Ohnaka et al. (2005) presented the resultsof the first mid-infrared interferometricobservations of the Mira variable RR Scowith MIDI, coordinated with K-band obser-vations using VINCI. The MIDI instrumenthas made it possible to measure the wave-length dependence of the visibility between8 and 13 µm with a spectral resolution of ~ 30 (Figure 8, panel a). In this spectral re-gion, a huge number of molecular lines dueto H2O and SiO are located, which allows oneto directly study the parameters of the mo-lecular gas in the outer atmosphere. Figure 8(panel b) shows the obtained wavelengthdependence of the angular diameter of

Wittkowski M. et al., Recent Astrophysical Results from the VLTI


RR Sco. The angular diameter of 15–24 masbetween 8 and 13 µm is significantly largerthan the K-band diameter of 10.2 mas meas-ured using VINCI, only three weeks after theMIDI observations. This wavelength depend-ence of the observed angular diameter of RRSco can be interpreted in terms of the pres-ence of a warm molecular envelope and anexpanding dust shell. Ohnaka et al. (2005)derived physical parameters of these compo-nents using the N-band visibilities and thespectrum obtained with MIDI together withthe K-band visibility measured with VINCIas observational constraints. Their model cal-culations show that optically thick emissionfrom the H2O + SiO envelope extending to ~2.3 stellar radii with a temperature of 1400 Kmakes the apparent mid-infrared diametermuch larger than the near-infrared angulardiameter. The increase of the angular diame-ter longward of 10 µm can be explained bythe presence of an optically thin dust shellconsisting of corundum and silicate with aninner radius at 7–8 stellar radii.

Additional insights into the conditions ofthe molecular gas in the outer atmospheresand circumstellar environment of AGB starscan be achieved by measuring and mappingthe maser radiation that some of these mole-cules emit. Boboltz & Wittkowski (2005)conducted concurrent observations of theMira variable S Ori as part of a program in-tended to exploit the power of long baselineinterferometry at infrared and radio wave-lengths to study the photospheres and nearbycircumstellar envelopes of evolved stars. Fig-ure 9 shows the results of the first-evercoordinated observations between NRAO’sVLBA (Very Long Baseline Array) andESO’s VLTI. The VLBA was used to observethe 43 GHz SiO maser emission (represent-ed by the circles color-coded in bins of radi-al velocity) concurrent with near-infrared K-band VINCI observations of the stellarphotosphere (represented by the red disc inthe centre of the distribution). The SiO maserswere found to lie at a distance of about 1.7 stellar radii or 1.5 AU. With concurrentobservations such as these, parameters of thecircumstellar gas, as traced by the SiO ma-sers, can be related to the star itself at a par-ticular phase in its pulsation cycle. Suchmeasurements can be used to constrain mod-els of stellar pulsation, envelope chemistry,maser generation, and stellar evolution.

Fundamental parameters, most important-ly radii and effective temperatures, of thephotospheres of the central cool giant starsthemselves have been frequently obtainedwith interferometric and other high angularresolution techniques, thanks to the favorablebrightness and size of these stars. Richichiand Roccatagliata (2005) recently conductednew accurate measurements of the angulardiameter of the cool giant Aldebaran usingnear-infrared interferometric and lunar oc-cultation techniques. They discuss explana-

tions for the elusiveness of a common angu-lar diameter value which includes temporalvariations as well as a dependence of the stel-lar diameter on wavelength.

Through the direct measurement of thecentre-to-limb intensity variation (CLV)across stellar discs and their close environ-ments, interferometry also probes the verti-cal temperature profile, the chemical com-position, and horizontal inhomogeneities.Limb-darkening studies have already beenaccomplished for a relatively small numberof star using different interferometric facili-ties. Wittkowski et al. (2004) presented thefirst limb-darkening observation that was ob-tained with the VLTI. Using the VINCI in-

strument, they measured K-band visibilitiesof the M4 giant ψ Phe in the first and secondlobe of the visibility function. These obser-vations were found to be consistent with pre-dictions by PHOENIX and ATLAS modelatmospheres, the parameters for which wereconstrained by comparison to available spec-trophotometry and theoretical stellar evolu-tionary tracks.

For cool pulsating Mira stars, the CLVs areexpected to be more complex than for non-pulsating M giants due to the effects of molec-ular layers close to the continuum-forminglayers. Broad-band CLVs may appear asGaussian-shaped or multi-component func-tions. Woodruff et al. (2004) and Fedele et al.

Figure 8: MIDI observations of the Mira variable RR Sco. Comparison between the visibilityobserved with MIDI (a), uniform-disc diameter (b),and spectrum (c) and those predicted by the best-fit model for RR Sco consisting of a warm H2O + SiO envelope and an optically thin dust shell.a: The diamonds and triangles represent the visibil-ities observed with projected baseline lengths of99.9 m and 73.7 m, respectively, while the corre-sponding predicted visibilities are represented withthe solid and dashed lines, respectively. b: The filled diamonds represent the observed uni-form-disc diameters, while the solid line representsthe model prediction. The dotted line representsthe continuum angular diameter estimated from theVINCI observation. c: The filled circles represent the calibrated spec-trum of RR Sco obtained from the MIDI obser-vations. The solid line represents the spectrum pre-dicted from the best-fit model. From Ohnaka et al.(2005).

Figure 9: First-ever coordinated ob-servations between ESO’s VLTI and NRAO’s VLBA: SiO maser emis-sions toward the Mira variable S Orimeasured with the VLBA, togetherwith the near-infrared diameter meas-ured quasi simultaneously with theVLTI (red stellar disc). From Boboltz & Wittkowski (2005).

41© ESO – March 2005Wittkowski M. et al., Recent Astrophysical Results from the VLTI

(2005) presented K-band VINCI observationsof the prototype Mira stars o Cet and R Leo,respectively. These measurements at post-maximum stellar phases indicate K-bandCLVs which are clearly different from a uni-form disc profile already in the first lobe ofthe visibility function. The measured visibil-ity values were found to be consistent withpredictions by recent self-excited dynamicMira model atmospheres (Ireland et al. 2004and references therein; Scholz & Wood 2004,private communication) that include molec-ular shells close to continuum-forming lay-ers. Figure 10 shows as an example the com-parison of the VINCI observation of the Mirastars R Leo (Fedele et al. 2005) with the pre-dictions by the dynamic model atmospheresmentioned above.

POST-AGB STARS

Further steps towards our better understand-ing of the stellar mass loss process are inter-ferometric measurements of post-AGB stars.One of the basic unknowns in the study oflate-type stars is the mechanism by whichspherically symmetric AGB stars evolve toform axisymmetric planetary nebulae (PNe).Antoniucci et al. (2005) presented the first de-tection of the envelope which surrounds thepost-AGB binary source HR 4049 by K-bandVINCI observations. They report on a phys-ical size of the envelope in the near-infraredK-band of about 15 AU (Gaussian FWHM).These measurements provide information onthe geometry of the emitting region and covera range of position angles of about 60 deg.They show that there is only a slight varia-tion of the size with position angle coveredwithin this range. These observations are,thus, consistent with a spherical envelope atthis distance from the stellar source, while anasymmetric envelope cannot be completelyruled out due to the limitation in azimuthrange, spatial frequency, and wavelengthrange. Further investigations using the near-infrared instrument AMBER can reveal thegeometry of this near-infrared component inmore detail, and MIDI observations can addinformation on cooler dust at larger distancesfrom the stellar surface.

Chesneau et al. (2005c) very recently pre-sented MIDI observations of the circumstel-lar environment of the massive OH/IR starOH26.5+0.6. This source is assumed to be atthe tip of the AGB phase, and to have enteredthe superwind stage only about 200 years ago.The emission of the dusty envelope, appearedto be resolved by a single UT. The MIDI ac-quisition image taken at 8.7 µm exhibitedclearly an asymmetry of the envelope of thisstar. In interferometric mode, no fringes weredetected. The authors argue that this failureto detect fringes, caused by an over-resolvedsize of the source at this wavelengh range,gives additional constraints on the opacitiesof the inner regions of the dust shell or closevicinity of the star.

CEPHEID PULSATIONS AND

DISTANCE ESTIMATES

For almost a century, Cepheid variable starshave occupied a central role in distance deter-minations. This is thanks to the existence of the Period-Luminosity (P – L) relation M = a log P + b which relates the logarithmof the variability period P of a Cepheid to itsabsolute mean magnitude M. This relation isthe basis of the extragalactic distance scale,but its calibration is still uncertain at a ∆M = ± 0.10 mag level. This uncertainty canbe overcome by an independent estimate ofthe distance to a sufficient number of nearbyCepheids. Trigonometric parallaxes are gen-erally too uncertain to provide strong enoughconstraints, and an alternative is provided bythe Baade-Wesselink (BW) method. Its basicprinciple is to compare the linear and angu-lar size variation of a pulsating star, in orderto derive its distance through a simple divi-sion. Interferometry allows one to measuredirectly the angular diameter variation dur-ing the pulsation cycle, while the linear sizevariation can be obtained by high resolutionspectroscopy (through the integration of theradial velocity curve). As described and re-viewed in detail by Kervella et al. (ESO Mes-senger, 2004), the VINCI observations of 7 southern Cepheids (Kervella et al. 2004c)allowed the calibration of the zero point ofthe P – L relation using interferometric BWdistance measurements. The resulting zeropoint value is identical to the one obtainedusing a large number of Hipparcos paral-laxes by Lanoix et al. (1999). This encourag-ing result strengthens confidence that nolarge bias is present on the P – L calibration(Kervella et al. 2004b). Additionally, theVINCI measurements provided new calibra-tions of the Cepheid Period-Radius and Sur-face brightness-Color relations (Kervella etal. 2004a), two important relations to con-strain theoretical models of Cepheids.

VERY LOW MASS STARS AND STELLAR

EVOLUTION NEAR THE MAIN SEQUENCE

The VLTI was used to precisely measure fun-damental parameters of very low mass dwarfstars and of stars close to the main sequence(MS). During evolution after the main se-quence, up to the subgiant and red giantstages, the stellar diameter increases enor-mously, up to several hundred times that ofthe present Sun. Even while the star is stillclose the the original MS, it inflates slowly.This means that the size of a given MS star islinked to its age. When combined with a the-oretical stellar evolution model, a directmeasurement of the size of a star thereforeprovides an estimate of its age.

Until recently, the mass-radius relation forvery low mass stars was poorly constrained.The VINCI observations of four nearby M dwarfs using the 8 m telescopes UT1 andUT3 allowed the measurement of accurateangular sizes, and improved our knowledgeof these faint stars (Ségransan et al. 2003). Inparticular, our nearest neighbour Proxima(M5.5V) was measured for the first time, witha very small radius of 0.145 RA, only slight-ly bigger than Jupiter (0.103 RA).

The angular diameters of a number ofnearby dwarf stars were measured with highprecision using VINCI, with the goal to estab-lish their age based on numerical models oftheir evolution. The addition of the diameteras a constraint reduces dramatically the un-certainty of the evolutionary state of the star.For instance, the two components of our near-est neighbor, the binary star α Centauri, wereresolved for the first time using VINCI(Kervella et al. 2003b). The 0.2% accuracyof the angular diameter measurement of α Cen A is among the highest precisions everachieved by interferometry. The numericalmodeling of α Cen A and B and constraintsfrom observed asteroseismic frequenciesavailable in the literature reproduces the di-

Figure 10: VINCI observations of theK-band intensity profile of the Miravariable R Leo and comparison todynamic model atmosphere predic-tions (Ireland et al. 2004 and refer-ences therein; Scholz & Wood, pri-vate communication). The top panelshows the CLV prediction by modelP22 (green lines: monochromaticCLVs; red line: filter-averaged CLV),and the bottom panel shows the cor-responding visibility curve (red line)compared to the measured R Leovisibility values. Also indicated in thetop panel are the mean positions ofthe SiO maser shells observed byCotton et al. (2004). Based on Fedeleet al. (2005).

2001, using six different siderostat and fivedifferent UT baselines. Based on angular di-ameters of these stars that have previouslybeen available in the literature, the authorsalso discuss the evolution of the measuredVLTI transfer function for the period 2001 tomid 2004. Work in progress aims at calculat-ing a global VLTI-based solution that willdetermine uniquely new diameters of a sub-set of about 50 chosen calibration stars byminimizing the scatter of the residuals foreach VLTI night under consideration.

Also, over the past two years, 16 new an-gular diameter measurements of nearby MSdwarfs and subgiants were obtained with theVINCI instrument, with the goal to calibratethe specific surface brightness-colour (SB)relations of these stars with high accuracy(Kervella et al. 2004e). The smallest disper-sions are obtained for the visible-infraredbased relations, in particular those based onthe B and K magnitudes. These relationsallow the prediction of very accurate (within1%) angular diameters of candidate calibra-tion stars for long baseline interferometry.

The MIDI consortium has established alist of calibration stars for the MIDI instru-ment based on spectro-photometric observa-tions of candidate stars and fitting of the datato atmosphere models (B. Stecklum, ESOcalibrator workshop 2003; van Boekel et al.,submitted).

ACKNOWLEDGEMENTSThe figures that appeared earlier in other journalswere reproduced by kind permission of theirauthors and the journals in which they originallyappeared. We are grateful for valuable discussionson the topics of this review with several colleagues.

PUBLICATIONS IN 2003 BASED ON VLTI DATA3

Domiciano de Souza, A., Kervella, P., Jankov, S.etal. 2003, A&A, 407, L47 (VINCI)

Kervella, P., Thévenin, F., Morel, P., Bordé, P. &Di Folco, E. 2003a, A&A, 408, 681 (VINCI)

Kervella, P., Thévenin, F., Ségransan, D., et al.2003b, A&A, 404, 1087 (VINCI)

Pijpers, F. P., Teixeira, T. C., Garcia, P. J. et al. 2003,A&A, 406, L15 (VINCI)

Ségransan, D., Kervella, P., Forveille, T., &Queloz, D. 2003, A&A, 397, L5 (VINCI)

van Boekel, R., Kervella, P., Schöller, M. et al.2003, A&A, 410, L37 (VINCI)

PUBLICATIONS IN 2004 BASED ON VLTI DATA3

Di Folco, E., Thévenin, F., Kervella, P. et al. 2004,A&A, 426, 601 (VINCI)

Jaffe, W., Meisenheimer, K., Röttgering, H. et al.2004, Nature, 429, 47 (MIDI)

Kervella, P., Bersier, D., Mourard, D. et al. 2004a,A&A, 428, 587 (VINCI)

Kervella, P., Bersier, D., Mourard, D., Nardetto, N.& Coudé du Foresto, V. 2004b, A&A, 423, 327(VINCI)

Kervella, P., Nardetto, N., Bersier, D., Mourard, D. & Coudé du Foresto, V. 2004c, A&A, 416, 941(VINCI)

Kervella, P., Ségransan, D. & Coudé du Foresto, V.2004d, A&A, 425, 1161 (VINCI)

Kervella, P., Thévenin, F. & Di Folco, E. 2004e,A&A, 426, 297 (VINCI)

Kervella, P., Thévenin, F., Morel, P. et al. 2004f,A&A, 413, 251 (VINCI)

Kervella, P., Fouqué, P., Storm, J. et al. 2004g, ApJ,604, L113 (VINCI)

Le Bouquin, J. B., Rousselet-Perraut, K., Kern, P.et al. 2004, A&A, 424, 719 (VINCI)

Leinert, Ch., van Boekel, R., Waters, L. B. F. M. etal. 2004, A&A, 423, 537 (MIDI)

Wittkowski, M., Aufdenberg, J. P. & Kervella, P.2004a, A&A, 413, 711 (VINCI)

Wittkowski, M., Kervella, P., Arsenault, R. et al.2004b, A&A, 418, L39 (VINCI)

Woodruff, H. C., Eberhardt, M., Driebe, T. et al.2004, A&A, 421, 703 (VINCI)

van Boekel, R., Min, M., Leinert, C. et al. 2004,Nature, 432, 479 (MIDI)

PUBLICATIONS IN 2005 (SO FAR) BASED ON

VLTI DATA4

Antoniucci, S., Paresce, F. & Wittkowski, M. 2005,A&A, 429, L1 (VINCI)

Boboltz, D. A. & Wittkowski M. 2005, ApJ, 618,953 (VINCI)

Chesneau, O., Meilland, P., Stee, P. et al. 2005a, A&A, in press (astro-ph/0501162) (MIDI)

Chesneau, O., Min, M., Herbst, T. et al. 2005b,A&A, in press (astro-ph/0501159) (MIDI)

Chesneau, O., Verhoelst, T., Lopez, B. et al. 2005c, A&A, in press (astro-ph/0501187) (MIDI)

Davis, J., Richichi, A., Ballester, P. et al. 2005, AN,326, 25 (VINCI)

Fedele, D., Wittkowski, M., Paresce, F., et al. 2005,A&A, 431, 1019 (VINCI)Ohnaka, K., Bergeat, J., Driebe, T. et al. 2005,

A&A, 429, 1057: (MIDI,VINCI)Richichi, A. & Percheron I. 2005, A&A, in press

(astro-ph/0501532) (VINCI)Richichi, A. & Roccatagliata, V. 2005, A&A,

433, 305 (VINCI)Thévenin, F., Kervella, P., Pichon, B. et al. 2005,

A&A, in press (astro-ph/0501420) (VINCI)

OTHER REFERENCESCotton, W. D., Mennesson, B., Diamond, P. J. et al.

2004, A&A, 414, 275Gallimore, J. F., Baum, S. A. & O’Dea, C. P. 2004,

ApJ, 613, 794Glindemann, A., Albertsen, M., Andolfato, L. et al.

2004, Proc. SPIE 5491, 447Ireland, M. J., Scholz, M. & Wood, P. R. 2004

MNRAS, 352, 318Kervella, P., Bersier, D., Nardetto, N. et al. 2004,

The Messenger, 117, 53Lanoix, P., Paturel, G. & Garnier, R. 1999, MNRAS,

308, 969Richichi, A. & Paresce, F. 2003, The Messenger,

114, 26Richichi, A., Percheron, I. & Khristoforova, M.

2005, A&A, 431, 773Wittkowski, M., Balega, Y., Beckert, T. et al. 1998,

A&A, 329, L45

3 From the ESO telescope bibliography(archive.eso.org/wdb/wdb/eso/publications/form)


ameters obtained by VLTI within their verysmall error bars.

The VINCI measurements of the brightstars Sirius A (Kervella et al. 2003a) andProcyon A (Kervella et al. 2004f) allowed theestimate of a very young age of 200–250 Myrfor the dwarf Sirius, and an older age of 2.3 Gyr for the subgiant Procyon. Figure 11shows as an example three evolutionary trackmodels of Procyon. The interferometric con-straint reduces considerably the uncertaintydomain in the HR diagram, especially whencoupled with asteroseismic observations.The age estimates based on the modeling ofthe interferometric data are confirmed by thecooling ages of the two white dwarfs that orbitSirius A and Procyon A.

Similar modeling studies were also con-ducted for the debris disc stars α PsA, β Pic,ε Eri and τ Cet (Di Folco et al. 2004). Thesestars are MS or pre-main sequence stars withspectral type A to K presenting a far infraredexcess, which has been commonly identifiedwith circumstellar dust in optically thin discs.These discs are in turn often associated with planetary systems presumably alreadyformed. The determination of the stellar agesis an essential information to study the evo-lution of the discs.

DIAMETER ESTIMATES OF

INTERFEROMETRIC CALIBRATION STARS

Interferometric measurements require themonitoring of the interferometric transferfunction, which in turn is derived from obser-vations of interferometric calibration stars.Usually those stars are selected as calibrationstars that do not show any peculiar character-istics, and that have a well-known angulardiameter and are as little resolved as possi-ble. The need for well-known interferomet-ric calibration stars has stimulated work aim-ing at measuring angular diameters of suchcandidates as well as establishing surfacebrightness-colour relations that allow theestimate of angular diameters based onknown brightness and colour.

Richichi & Percheron (2005) recently de-scribed a massive ongoing effort, that wasstarted in 2001, to accumulate observationson calibration stars with the aim of develop-ing a VLTI-based system of high accuracycalibration stars. Observations of 191 calibra-tion star candidates have been conducted withthe VLTI and the near-infrared K-band com-missioning instrument VINCI since March

Figure 11: Evolutionary tracks in the Hertzsprung-Russel diagram of three models of Procyon A. Thedashed rectangle delimits the classical uncertaintydomain for luminosity and effective temperature,while the hatched area delimits the much reduceduncertainty when considering the interferometricradius. Model a is the most probable and indicatesan age of 2.3 Gyr. From Kervella et al. (2004f).

4 Publications that have appeared or are in pressand that qualify to be included in the ESO tele-scope bibliography (as of 15 February 2005). Notethat this list may be incomplete.


TThe nature of star formation andevolution close to a super-mas-sive black hole is of broad as-trophysical interest. Due to itsproximity (~ 8 kpc), the centre

of our galaxy (GC) offers a unique variety ofexperiments and observations, which growstogether with the technical progress and thecommissioning of new instruments. At pres-ent, the angular resolution of the VLTI al-ready allows us to resolve the dust envelopesof some stars in the GC. A structural analysison the scale of tens to hundreds of AUs opensthe way for a detailed study of stellar proper-ties, as well as of the interaction between astar and the GC environment.

The unusually large number of massive,young stars in the stellar cluster at the GC(e.g. Genzel et al. 2003, Eckart et al. 2004,Moultaka et al. 2004) is indicative of an activestar formation history despite the tidal forcesexerted by the gravitational potential of thecentral black hole. The presence of numerousstars in short-lived phases of their develop-ment, such as dust producing Wolf-Rayet(WR) stars, indicates that the most recent starformation episode took place not more thana few million years ago. IRS 3 with its 1–2arcsec extended mid infrared (MIR) excess isone of the most prominent of these sources(Viehmann et al. 2005; Moultaka et al. 2004).

Why is this source a good starting pointfor VLTI observations of the GC region withMIDI, the MID-infrared Instrument for theVLT Interferometer? Pott et al. (2004) re-

viewed the technical aspects of VLTI-GCobservations, which are ideally suited tostudy the capabilities of the new instrumentsclose to the system limits under normal ob-serving conditions. Here we focus mainly onastrophysical aspects.

The nature of IRS 3 is not yet identifiedunambiguously. In the late 1970s, it wasargued that IRS 3 is a dust-enshrouded super-giant with a compact circumstellar dust shell.IRS 3 was found to be the most compact and(together with IRS 7) hottest MIR source (T ~ 400 K) in the central cluster, with totalintegrated flux densities of about 30 Jy at 8 µm to 12 µm.

Given its high luminosity of ~ 5 · 104 LA

IRS 3 may in fact be a star at the very tip ofthe Asymptotic Giant Branch (AGB). Thesemost luminous dust-enshrouded AGB starswill stay at a high luminosity during theirentire mass loss phase. Their mass-loss rates,derived from observations, span a range fromabout 10–7 to 10–3 MAyr–1 (van Loon et al.1999) with wind velocities of the order of 10–20 km/s. In general the more luminousand cooler stars are found to reach highermass-loss rates. This is in agreement withmodel calculations. For a synthetic sample ofmore than 5 000 brighter tip-AGB stars a col-lective mass-loss rate of 5.0 × 10–4MAyr–1

was found. Of these, 20 are carbon-rich su-per-giants with a large IR excess and a mass-loss rate well in excess of 10–6MAyr–1, includ-ing 10 dust-enshrouded, extreme tip-AGBstars seen in their short-lived (~ 30 000 yrs)

super-wind phase with a mass loss of >10–5MAyr–1.They produce about 50% of thecollective mass-loss of the whole sample.

Arecent identification of a carbon-rich WRstar of type WC5/6 as a near infrared coun-terpart of IRS 3, based on the detection of a2.11 µm He I/C III line (Horrobin et al. 2004),is probably applicable to a K ~ 15 faint star ~ 120 mas east of the bright source. However,given the fact that most other dust enshroud-ed sources in the central stellar cluster havebeen associated with hot and luminous youngstars, an identification of IRS 3 with a mas-sive WR star in its dust forming phase can-not be fully excluded either. Extensive massloss associated with bright continuum emis-sion takes place in the WC stage. Products ofhelium burning are dredged up to the sur-face, enhancing the carbon and further de-pleting the hydrogen abundance.

As shown by Viehmann et al. (2005), thedust shell of IRS 3 is interacting with the GCISM. They find the photocentre of IRS 3 inthe ISAAC M-band image shifted by ~ 160mas to the NW with respect to the L-bandimage.About 1)to the southeast of IRS 3 high-pass filtered L- and M-band NAOS/CONICAimages show a sharp interaction zone of theouter part of the dust shell with the wind aris-ing from the IRS 16 cluster of hot, massiveHelium stars.

We designed a VLTI experiment withMIDI (N-band, 8–12 µm) to investigate thedust shell of IRS 3. The lower spectroscopicresolution used (R = 30) offers dispersed

VLVLTI OTI OBSERBSERVVAATIONSTIONS OFOF IRS 3:IRS 3:TTHEHE BRIGHTESTBRIGHTEST COMPCOMPACTACT MIR MIR SOURCESOURCE AATT THETHE

GGALACTICALACTIC CCENTREENTRE

THE DUST ENSHROUDED STAR IRS 3 IN THE CENTRAL LIGHT YEAR OF OUR GALAXY WAS PARTIALLY RESOLVED

IN A RECENT VLTI EXPERIMENT. THIS OBSERVATION IS THE FIRST STEP IN INVESTIGATING BOTH IRS 3 IN

PARTICULAR AND THE STELLAR POPULATION OF THE GALACTIC CENTRE IN GENERAL WITH THE VLTI AT THE

HIGHEST ANGULAR RESOLUTION. WE OUTLINE WHAT SCIENTIFIC ISSUES CAN BE ADDRESSED BY A COMPLETE

MIDI DATASET ON IRS 3 IN THE MID-INFRARED.

JJÖRGÖRG-U-UWEWE PPOTTOTT 11,2,2,, AANDREASNDREAS EECKARCKARTT 11,, AANDREASNDREAS GGLINDEMANNLINDEMANN22,, TTHOMASHOMAS VVIEHMANNIEHMANN11,,RRAINERAINER SSCHÖDELCHÖDEL11, C, CHRISTIANHRISTIAN SSTRAUBMEIERTRAUBMEIER11, C, CHRISTHRISTOPHOPH LLEINEREINERTT 33, M, MARKUSARKUS FFELDTELDT 33,,

RREINHARDEINHARD GGENZELENZEL 44 ANDAND MMASSIMOASSIMO RROBBEROBBERTTOO 55

1UNIVERSITY OF COLOGNE, COLOGNE, GERMANY2EUROPEAN SOUTHERN OBSERVATORY, GARCHING, GERMANY

3MAX-PLANCK-INSTITUT FÜR ASTRONOMIE, HEIDELBERG, GERMANY4MAX-PLANCK-INSTITUT FÜR EXTRATERRESTRISCHE PHYSIK, GARCHING, GERMANY

5ESA-SPACE TELESCOPE SCIENCE INSTITUTE, BALTIMORE, MD, USA


visibility data over the entire N-band, as wellas a spectrum of the uncorrelated flux densi-ty. The first VLTI detection of a star in the GCwas achieved in June 2004: We partially re-solved IRS 3 with the VLTI using MIDI onthe 47 m UT2-UT3 baseline (see Figures 1and 2).

It was found that ~ 25% of the flux densi-ty of IRS 3 is concentrated in a compact (i.e. unresolved) component with a size of ≤ 40 mas (i.e. ≤ 300 AU). This agrees withthe interpretation that IRS 3 is a luminouscompact object in an intensive dust formingphase. In general, the visibility amplitude wasfound to increase with wavelength by ~ 0.05.Although the uncertainty of a single visibili-ty value (5–10%) seems to be too large to un-ambiguously identify such a trend, the erroron the slope (i.e. wavelength dependent vari-ation) of a visibility dataset over the entire N-band is of the order of 1% only. This trendindicates that the compact portion of the IRS 3 dust shell is extended and only partial-ly resolved on the UT2-UT3 baseline. We alsofind indications for a narrower width of the9.3 µm silicate line towards the centre, in-dicating the presence of fresh unprocessedsmall grains closer to the central star in IRS 3 (van Boekel et al., 2003).

In addition, the remaining six of the sevenbrightest (N-band) MIR excess sources in theGC were observed (IRS 1W, 2, 8, 9, 10, 13)with the same instrumental setup. Most ofthem are located in the Northern Arm of theISM or associated with the mini-spiral ofionised gas and warm dust (Figure 1). Theyappear to be hot stars with strong, fast windsthat create bow shocks as they plough throughthe gas and dust of the mini-spiral. The MIRemission associated with theses sourcesarises most probably from these bow shocksthat can be resolved at 2 µm.

All of these six sources are fully resolvedat a 2 Jy flux level, i.e. only upper limits onthe visibility limits could be estimated withthe baseline used. Therefore, IRS 3 is not onlythe hottest but also the most compact brightsource of MIR emission within the centralstellar cluster.

We will conduct further VLTI/MIDI ob-servations of IRS 3 to enlarge the uv-planecoverage (Figure 3). The foreseen baselinesUT3-UT4 (62 m) and UT1-UT4 (130 m) are complementary to UT2-UT3 in terms oflength and orientation. The final dataset willcover an angular resolution of about 13 masat the longest baseline up to 100 mas. Theseobservations will be ideally suited to provideinformation about the radial structure andsymmetry of the correlated flux, about theinner edge of the dust shell, as well as abouta possible binary character of IRS 3. Colli-sions of winds in binary systems may supportdust formation through density increase andrapid cooling of the material.

A bright compact source like IRS 3 is alsotechnically essential for future VLTI phase-

reference experiments at 10 µm in order to investigate other nearby sources, e.g. the Sgr A* black hole. For this purpose it isimportant to know the strength and compact-ness of IRS 3 on the longest baselines.

REFERENCESvan Boekel, R., Waters, L. B. F. M., Dominik, C.

et al. 2003, A&A, 400, L21Eckart, A., Moultaka, J., Viehmann, T. et al. 2004,

ApJ 602, 760Genzel, R., Schödel, R., Ott, T. et al. 2003, ApJ 594,

812

Horrobin, M., Eisenhauer, F., Tecza, M. et al. 2004,AN, 325, 88

Moultaka, J., Eckart, A., Viehmann, T. et al. 2004,A&A 425, 529

Pott, J.-U., Glindemann, A., Eckart, A. et al.2004,SPIE, 5491, 126

Schödel, R., Ott, T., Genzel, R. et al. 2002, Nature,419, 694

van Loon, J. Th.; Groenewegen, M. A. T.; de Koter,A. et al. 1999, A&A, 351, 559

Viehmann, T., Eckart, A. Schödel et al. 2005,accepted by A&A, astro-ph/0411798

Figure 1: VISIR MIRimages with the ob-served targets in-dicated and an insetdemonstrating thescale on which thecurrent UT2-UT3VLTI/MIDI data detect-ed a compact sourcewith a visibility of about 25%.

IRS 8

IRS 9

IRS 2

IRS 13

IRS 7 IRS 3

IRS 10

IRS 1EW

N

5 arcsec 5 arcsec

1 arcsec

VISIR 8.6 micron

VISIR 12.8 micronVLTI/MIDI

2.00.1

0.2

0.3

0.4

0.5

2.5 3.0 3.5 4.0 4.5 5.0

visi

bili

tyam

plit

ude

uvradius in 106 (”mega-lambda”)

-6-6

-4

0

2

4

6

-4 -2 0 2 4 6

-2

106 (”mega-lambda”)

106

(”m

ega-

lam

bd

a”)

Figure 2: Median 8–12 µm visibility ofIRS 3 as a function ofprojected baselinelength. Currently theuncertainty of thevisibilities is above allaffected by the in-strument calibration.Therefore the errors are given by the stan-dard deviation of theinstrument calibrationover the entire night.

Figure 3: The uv-cover-age of the observationsin P73 is shown. Thedotted line indicatesthe change of projectedbaseline length due toearth rotation. Whereasthe earth rotation curve is calculated at a cen-tral wavelength of10.34 µm the solid linesare showing the uv-coverage of each dis-persed, calibrated visi-bility dataset.


TTHE EUROPEAN SOUTHERN

Observatory (ESO) and theSpace Telescope Science Insti-tute (STScI) have developedscience metrics tools to mea-

sure and monitor the scientific success oftheir telescopes. Two recent papers outlinethe methodology used to produce the metricsand evaluate the results: “Metrics to MeasureESO’s Scientific Success,” by Leibundgut,Grothkopf & Treumann (2003) and “HubbleSpace Telescope Science Metrics”, byMeylan, Madrid & Macchetto (2004).

In an attempt to put the results into a broad-er context, we present here a comparisonbetween the individual science metrics ofHST (Hubble Space Telescope) and VLT(Very Large Telescope). The two facilitieshave rather different, yet often complemen-tary, capabilities. They also serve a similarcommunity of optical and near-IR astrono-mers. It is generally acknowledged that theyhave leading roles in astronomical research.Here we try to assess the relative roles andthe scientific impact of the two observatories.

The Hubble Space Telescope, launched inApril 1990, is the product of collaboration be-tween NASA and the European Space Agen-cy. During its lifetime HST has had nine ex-tremely successful instruments, six of whichare by now decommissioned. Unique capa-bilities of HST include the access to UVwavelengths, the unmatched large-field, highspatial resolution and the low background atnear-infrared wavelengths.

The Very Large Telescope is one of themajor ground-based observatories, with four8 m unit telescopes and the capability of com-bining them (and smaller auxiliary tele-scopes) for interferometry. In operation since1999, the VLT has grown to include all fourunit telescopes and ten instruments are cur-rently offered to the community. The instru-ment suite covers most wavelengths accessi-

ble from the ground, adaptive optics instru-ments for small-field high spatial-resolutionimaging, high-resolution spectroscopy andmulti-object spectroscopy.

Methodologies for assessment of scien-tific impact have been developed by sev-eral authors for individual observatories (Meylan, Madrid & Macchetto 2003, 2004;Leibundgut, Grothkopf & Treumann 2003,Crabtree & Bryson, 2001), comparisons be-tween observatories (Trimble 1995, Benn &Sanchez 2001, Trimble, Zaich & Bosler2005) or generally the impact of journals orastronomical subfields (Abt papers, Schwarz& Kennicutt 2004). Almost all investigationsconverge on the basic data input for the met-rics: numbers of papers published, citationsto these papers, and impact measured throughhighly cited papers. Differences in the analy-ses mostly stem from different input data-bases and weighting of stated goals of theinvestigation, e.g. which audience is ad-dressed. For observatories, the prime interestis, of course, to evaluate how the data col-lected with their facilities contribute to sci-entific progress.

In the following we use the number of ref-ereed publications as one measure of theproductivity of the VLT, HST and other ob-servatories, which provided us with the ap-propriate input data. We then proceed toinvestigate the number of citations these pa-pers generate, which measures their impactwithin the scientific astronomical communi-ty. Further investigations try to find the high-impact papers based on observatory data.

As of the end of 2004, 120 refereed paperscombined data from HST and VLT. The com-plementarity of these two observatories is re-flected in these papers. We performed ananalysis to investigate what projects makeuse of both observatories.

Our study is based on databases that arecomplete in the sense that they collect all

papers that use data from the respective ob-servatory. The databases are maintained atESO and STScI separately and we made aneffort to homogenize them for the compari-son.

NUMBER OF PAPERS BASED ON HSTAND VLT DATA

In order to compare publication statisticsamong different observatories, it is essentialto assess the selection criteria applied by eachobservatory and to relate the different crite-ria to each other in a meaningful way. Thefirst ingredient is a complete list of papersbased on observatory data. One might thinkthat it would be easiest to rely on the men-tioning of facility usage by the authors, forinstance through acknowledgements in foot-notes or through the use of data set and facil-ities identifiers1. However, these linking ini-tiatives are voluntary for the authors. Theywill only help facilities to derive metricswhen they are applied widely and reliably.

Automated searches in the NASA Astro-physics Data System (ADS) do not lead tocomplete lists of papers. A2002 study at ESOshowed that retrieval of papers through ADSalone missed approx. 20% of relevant papers,while at the same time an equal number ofpapers was erroneously included (Grothkopf& Treumann 2003). Similar findings havebeen confirmed in various further trials. Themain reason for the unreliability of automat-ed searches is not so much the unavailabilityof full texts (ADS currently covers only title,abstract and keywords of recent papers), butrather the fact that retrieval tools are not capa-ble of interpreting the context in which searchterms appear. Thus, we still need human as-

1 Identifiers for data sets and facilities are current-ly being introduced by AAS journals. They aremeant to improve navigation between scientificpapers and online data, see for instance Eichhorn et al. (2003).

Oth

erA

stro

nom

ical

New

s

CCOMPOMPARISONARISON OFOF SSCIENCECIENCE MMETRICSETRICS

AMONGAMONG OOBSERBSERVVAATORIESTORIES

A COMPARISON OF VARIOUS SCIENTIFIC METRICS IS PRESENTED. IN PARTICULAR, WE EXAMINE THE IMPACT OF

THE HST AND VLT OBSERVATORIES THROUGH PUBLICATION RATES, CITATION RATES AND THE NUMBER OF

HIGHLY CITED PUBLICATIONS. BOTH OBSERVATORIES ARE MAJOR CONTRIBUTORS TO TODAY’S ASTROPHYSICS

RESEARCH ENDEAVOUR AND CURRENTLY RANK AMONG THE MOST SUCCESSFUL ASTRONOMICAL FACILITIES AS

MEASURED BY THE ABOVE CATEGORIES.

UUTTAA GGROTHKOPFROTHKOPF 11, B, BRUNORUNO LLEIBUNDGUTEIBUNDGUT 11,,DDUCCIOUCCIO MMACCHETTOACCHETTO 22, J, JUANUAN PP. M. MADRIDADRID 22, C, CLAUSLAUS LLEITHEREREITHERER 22

1EUROPEAN SOUTHERN OBSERVATORY, GARCHING, GERMANY2SPACE TELESCOPE SCIENCE INSTITUTE, BALTIMORE, MD, USA


sessment in order to distinguish between pa-pers that meet our selection criteria and thosethat mention search terms in contexts thatare irrelevant for our purpose.

At most large observatories, librarians ordedicated personnel screen the literature andidentify all articles based on data from the re-spective facilities. These databases are usedfor multiple purposes including, for example,the Annual Report’s section on scientific arti-cles published by the observatory. At ESOand STScI, we have achieved a high reliabil-ity through manual screening of print jour-nals in combination with searches using ADS,and created comprehensive databases.

The HST publication database containspapers since the special HST Letters issue ofthe Astrophysical Journal in March 1991; theESO telescope bibliography contains publi-cations from 1996 onwards for La Silla pa-pers and from 1999 onwards for VLT papers.

Our databases list refereed papers that useHST or VLT data pertaining to at least one ofthe following data categories: (i) PI/CoI(Principal Investigator/Co-Investigator) data,i.e. publications in which at least one of theauthors was a participant of the original pro-posal; (ii) archival data, i.e. papers using datawhere none of the authors was a member ofthe original proposal group; and (iii) papersusing public data sets. We do not includepapers that merely cite results published inprevious articles. ESO excludes conferenceproceedings, even if published in refereedjournals. STScI, in contrast, includes confer-ence proceedings in refereed journals.

One would expect that the HST numbersof publications are somewhat enhanced com-pared to those of ESO due to this slight dif-ference. At the same time the citation ratesare possibly decreased as citations to confer-ence papers, even if refereed, are typicallylower than in the refereed literature (e.g.Schwarz & Kennicutt 2004). Despite thesedifferences, the statistics derived from thetwo databases are comparable, bearing inmind that the overall aim is to compare themin a broader context rather than on the basisof cross-checks among individual papers. Webelieve that the metrics we derived are objec-tive and reproducible.

At ESO, papers are categorized by tele-scopes, instruments and observing modes(visitor or service). It is also noted whetherthese are original observations or archivaldata. For all papers using VLT data, the pro-gramme IDs that generated the data areobtained either from the publication, from theobserving schedule or from the authors.

The ESO and STScI databases are public-ly available. The ESO bibliography can besearched using the web interface describedby Delmotte et al. on page 50 of this Messen-ger issue. In addition, but without search op-tions for specific telescopes or instruments,both databases are available through the ADS

Abstract Service by activating the filtersESO/Telescopes and HST, respectively. Ac-tive links to the corresponding archives areprovided for all papers using HST and VLT.

Figure 1 gives the total numbers of papersper year for the VLT and the HST. For com-parison, we also include publication numbersfor the Keck Observatory and the Chandra X-ray Observatory. We obtained these valuesfrom the Keck Science Bibliography (http://www2.keck.hawaii.edu/library/keckpub00.html) and from the Chandra Data Archive(http://cxc.harvard.edu/cgi-gen/cda/bibli-ography; see also the article by Green andYukita 2004).

The VLT and the HST are both very pro-ductive telescopes: as of January 2005, HSThas provided data for a total of 4 634 refereedpapers (over 14 years), VLT for a total of 938(over six years). Both observatories, as wellas Chandra, profit from a large user commu-nity. The start-up years of the VLT are near-ly identical in numbers of papers publishedcompared to the early years of HST opera-tions showing a strong and regular increaseper year. In the case of the VLT, part of thisincrease is due to the continuous expansion

of the facility and the addition of new tele-scopes and instruments. The instrument com-plement increased from two instruments inthe first year to four for the second, third andfourth year of operations, with a furtherincrease to today’s complement of ten instru-ments.

The HST reached a publication rate ofabout 500 refereed papers per year, eightyears after launch. The impressive contribu-tion of the HST reflects the fact that it pro-vides unmatched spatial resolution imagingat ultraviolet through near-infrared wave-lengths as well as spectroscopic observationsat wavelengths that ground-based telescopescannot access. Furthermore, the HST hasbeen regularly refurbished with new genera-tions of instruments through successive serv-icing missions, maintaining state-of-the-arttechnology.

The publication rate for the VLT is stillincreasing at the rate HST was displayingduring its first eight years. As an aside we notehere that up to 2004 there are still more paperspublished per year based on data obtainedwith telescopes at La Silla than from the VLT(337 vs. 344 in 2004).

Figure 1: Number ofrefereed papers usingVLT, HST, Keck, andChandra data (as ofJanuary 2005).

ApJ/ApJS

AJ

A&A

MNRAS

PASP

Nature

Others

Total

HST (1991–2004)

# of papers

2 269

851

586

369

106

47

406

4 634

%

48.9

18.4

12.6

8.0

2.3

1.0

8.8

100.0

VLT (1996–2004)

# of papers

235

52

527

74

1

18

31

938

%

25.1

5.5

56.2

7.9

0.1

1.9

3.3

100.0

Table 1: Distribution (absolutenumbers and percentages) of ref-ereed HST and VLT papers pub-lished in the five core astronomyjournals as well as Nature (as of January 2005). Note that theseare the total numbers of paperspublished over the current lifetimeof the observatories, i.e. 14 yearsfor HST and six years for the VLT.


A direct comparison is difficult at this stage.The HST is a mature observatory with a largeuser community and has operated over morethan a decade. The VLT is still in its ramp-upphase. Neither the total numbers of papers northe numbers of papers per year provide a goodindication how the two observatories com-pare to each other. Figure 1 needs to be inter-preted rather carefully in this respect. It willhave to be seen what level the VLT will reachafter several years.

About 90% of the HST papers and 95%of the VLT papers that were published untilthe end of 2004 appeared in the five majorastronomical journals, i.e. AstrophysicalJournal and its Supplements (ApJ/ApJS),Astronomical Journal (AJ), Astronomy &Astrophysics (A&A), Monthly Notices of theRoyal Astronomical Society (MNRAS) andPublications of the Astronomical Society ofthe Pacific (PASP). Table 1 shows the ab-solute numbers and percentages of HST andVLT publications in these journals. For com-parison, the corresponding numbers for pub-lications in Nature are also included. Thetable reflects the partly different affiliationsof the user communities of the two observa-tories. ApJ continues to be the preferred jour-nal for most authors of HST-based papers. Onthe other hand, the ESO user communityprefers to publish in A&A in which it isexempt from page charges.

CITATIONS

In order to assess the scientific impact of ref-ereed papers based on HST and VLT data, weobtained the number of citations for eachpaper in our databases from ADS. We areaware that the set of citations provided byADS is not complete, mainly because someciting journals are not within the ADS data-base. However, with regard to the major ref-ereed astronomical journals ADS is a veryreliable source of citations. For many majorastrophysics journals, the coverage is evencomplete back to volume 1.

Recently, Trimble, Zaich & Bosler (2005)argued that the contribution to papers basedon several observatories should be appor-tioned and split according to the contributionsfrom each observatory. This is a reasonableapproach for investigations which aim to de-duce general trends in astronomy publishing.For observatories such an approach is notreally suitable as the weighting process (likethe one in Trimble et al.) would require re-sources that are not available at the moment.The critical information for observatories ishow often their data have made a contributionto the scientific literature. Hence we chose tocount every paper as a full contribution in ourdatabases and analysis, even if several tele-scopes provided data. This of course meansthat papers making use of VLT and HST datawill show up in both databases (see section“Papers based on VLT and HST data at thesame time”). For the high-impact papers (see

below) we actually do weight the contribu-tion by each observatory.

Figure 2 shows the total number of cita-tions for HST and VLT papers by year. Bothobservatories are contributing significantlyto astrophysics research and the publicationsbased on their data form an integral part ofthe astrophysical literature. The HST stillleads in the total citations (as it does in the number of papers published, cf. Figure 1).However, there have been significant changesover the past few years. The VLT has a steadi-ly increasing citation rate compared to otherobservatories, while Chandra has made a sub-stantial impact right from the start. The Keck

Figure 2 (above): Totalnumber of citations toVLT, HST, Keck andChandra publicationsby year (as of Decem-ber 2004).

citation rates remain constant at a high level.This is illustrated in Figure 3 which showsthe evolution of the numbers of citations rel-ative to HST. This figure removes the ‘histo-ry’ effect inherent in citation statistics andallows a direct comparison of the observato-ry statistics.

Average citation rates are a highly simpli-fied means of assessing the success of publi-cations. The distribution of citations perpaper is highly non-uniform and hence aver-ages and standard deviations are only crude,if at all appropriate, means for such an analy-sis. One should also keep in mind that thereare severe differences in the way papers are

Figure 3 (below): Num-ber of citations perpaper relative to HSTcitations (as of Decem-ber 2004).

Grothkopf U. et al., Comparison of Science Metrics among Observatories


cited per astronomical subfield and other sec-ondary effects. A very good analysis of suchadditional parameters is given by Schwarz &Kennicutt (2004). To offset these statisticaldistortions we decided to use the median,which is in general a much more robust sta-tistical indicator. Hence, Figure 4 shows themedian number of citations per paper. Thesteady increase of the VLT citations is an indi-cation of the non-static situation of these sta-tistics. At the moment, the VLT and the HSTare nearly identical considering the meancitations of papers. The differences might bea signature of the different age of the two ob-servatories, although a more detailed analysismight be warranted here.

We further analyzed the 100 most fre-quently cited papers based on VLT and HST data, respectively. Interesting trends areshown in Figure 5. Almost half of the mostcited papers based on VLT data (1999–2004)were published in A&A (43% of the top 100VLT papers, citations from ADS), reflectingthe preference of publishing VLT data in A&A(cf. Table 1). Many of the most-cited publi-cations appeared in ApJ (39%). The share ofhighly cited papers published in Nature (9%)is large compared to the overall percentageof VLT papers published in this journal (ap-prox. 2.0%, cf. Table 1). In fact, half of theVLT papers published in Nature made it intothe top 100.

Aclear majority of the top 100 HST papers(1991–2004) were published in ApJ/ApJS(67%); this fraction is even larger than thealready high percentage of all papers pub-lished in these journals as given in Table 1(approx. 50%). As with VLT, articles pub-lished in Nature are highly representedamong the most cited papers: 1% of all HSTarticles are published in this journal vs. 5%among the top 100.

HIGH-IMPACT PAPERS

The term High-Impact Papers (HIP) wascoined by the Institute for Scientific Infor-mation (ISI) (see http://www.isinet.com/rsg/hip/). It defines the 200 most cited papers fora given year. The numbers presented herewere obtained from the ADS by querying thedatabase for 200 papers, sorted by descend-ing citation count. The method we used isdescribed in greater detail in Meylan, Madrid& Macchetto (2004). These 200 most citedpublications could be refereed or unrefereed,observational or theoretical papers. For eachof them we read the full text, selected thosethat present observations, and attributed thecitations to the facility or facilities that con-tributed the data. In the case of usage of datafrom multiple observatories, citations are dis-tributed according to a weight correspondingto the amount of data provided by each facil-ity. As a result, each observatory received anormalized percentage of its contribution tohigh impact papers.

Figure 5 (below): Frac-tion of most cited VLT(left) and HST papers(right) published in vari-ous journals.

The citations for each year were obtained 16 months after the year ended, i.e., citationsfor the year 2001 were retrieved in April2003, for the year 2002 in April 2004 etc. Atthis time, the citations had not yet reachedtheir peak (typically after two years, seeMeylan, Madrid & Macchetto 2004). Tomake sure that the relative impact per facili-ty remains the same also over a longer time,we cross-checked the high impact paperspublished in 1999 and 2001 in December2004 (five and three years respectively after

publication) and found that only 3% of the100 most cited 1999 papers and 7% of the 100 most cited 2001 papers had changed.

In Figure 6 we present the evolution of thecontribution of selected facilities to theseHIP. The number of available facilities, bothground- based and in space, is increasing andmany of them contribute to high impact pa-pers. For instance Chandra and the VLT con-tributed to HIPs for the first time in 2000.Accordingly, the share of already existingfacilities becomes smaller.

Figure 4 (above): Medi-an number of citationsfor VLT and HST pa-pers as a function ofyears since publication(publication years2003–1999, as of Jan-uary 2005).


Figure 6: Contributionof selected space andground-based facilitiesto the total citations of the 200 most citedpapers per year. NOAOincludes papers basedon Kitt Peak Observa-tory and CTIO.

PAPERS BASED ON VLT AND HST DATA AT

THE SAME TIME

HST and VLT data are increasingly combinedin the same paper. The collation of multi-wavelength datasets and the comparison ofcomplementary observations across the elec-tromagnetic spectrum have significantly im-proved the physical picture that astronomersobtain from celestial objects. This leads topublications based on observations comingfrom several sources. The interdependencebetween ground- and space-based observa-tions is steadily increasing. This trend wasconfirmed by Trimble, Zaich & Bosler (2005)and reflects a change in the way astrophysi-cal research is being conducted. This can fur-ther be seen in the science cases for futurefacilities, which typically explore the expect-ed synergies with other projects. Up throughthe end of 2004, 120 refereed papers used datafrom both observatories. Many authors com-bine the unique capabilities of the two facil-ities to obtain new astronomical knowledge.The most common cases we encounteredwere papers where high-resolution VLTspec-tra are combined with HST archival images;several papers presented follow-up VLTspectroscopy of objects previously observedwith HST. This pattern had previously alsobeen observed between HST and Keck. Thedeep fields are prime examples. In particularISAAC and the FORS instruments are usedin parallel with data obtained with WFPC2,ACS and NICMOS. These papers show clear-ly that ground-based telescopes and spacemissions do not compete, but support andcomplement each other and lead to higher sci-entific productivity through effective use ofcollaborations.

CONCLUSION

HST and VLT are amongst the most prolificobservatories and are fully competitive withother large facilities like Chandra and Keck.They also have taken the leadership role fromolder observatories, although the rate of pub-lication of data for example from La Silla tele-scopes still remains high.

After slightly more than five years of oper-ations, the VLT shows a similar developmentobserved in other successful observatories.Soon after First Light, publication statisticsstarted to rise steeply and are rapidly ap-proaching 1000 refereed publications by theend of 2004. These first five years are nearlyidentical to the development of HST duringits early years.

An equally positive trend can be noticedfor citations made to publications based onESO data. The mean number of citations perrefereed paper tends to be at the level of estab-lished observatories. The citation rate is stillrapidly increasing and is approaching theHST citation rate.

Astudy of the contributions to high impactpapers from some of the major telescopesavailable at present shows that the VLT con-tributes to a considerable fraction of them andthus occupied a place among the most impor-tant facilities only a few years after the startof its operations. Interesting differences canbe seen regarding where papers from the VLTand HST are published. While for the Euro-pean community, the prime user of the VLT,Astronomy & Astrophysics remains the jour-nal of choice, the majority of HST data arepublished in the Astrophysical Journal. Mostother journals are used with a comparable fre-quency.

HST continues to dominate the numbercount of published papers per year and shows

no sign of contributing less than in past yearsto the overall impact that these papers haveon astronomical research.

ACKNOWLEDGEMENTSThis article has made use of NASA’s Astrophys-ics Data System. We are grateful to AngelikaTreumann, who maintains the ESO telescope bib-liography and gave essential comments on the man-uscript. Our thanks also go to the STScI librarianSarah Stevens-Rayburn, who has made invaluablecontributions to the HST bibliography; Sarah alsokindly removed the non-native English speakers’flavour from the manuscript. We thank KarenLevay from MAST for her help in handling the database of the HST bibliography. Arnold Rotsfrom Harvard/Smithsonian Center for Astro-physics, as well as Peggi Kamisato from W. M.Keck Observatory kindly provided information onChandra and Keck publication and citation statis-tics.

REFERENCESAbt, H. A. 1981, PASP, 93, 207Abt, H. A. 1985, PASP, 97, 1050Benn, C. R. & Sanchez, S. F. 2001, PASP, 113, 385Delmotte, N. ,Treumann, A., Grothkopf U. et al.

2005, The Messenger, 119, 50Crabtree, D. R. & Bryson, E. 2001, JRASC, 95, 259Eichhorn, G., Accomazzi, A., Grant, C. S. et al.

2003, BAAS 35 #5, 20.04Green, P. J. & Yukita, M. 2004, Chandra News-

letter, issue 11, March 2004Grothkopf, U. & Treumann, A. 2003, LISA IV

proc., Corbin, B., Bryson E., Wolf, M. (eds.),Washington, DC: U. S. Naval Observatory, 193

Leibundgut, B., Grothkopf, U. & Treumann A.2003, The Messenger, 114, 46

Meylan, G., Madrid, J. P. & Macchetto, D. 2003,STScI Newsletter, 20, no. 2, 1

Meylan, G., Madrid, J. P. & Macchetto, D. 2004,PASP, 116, 790

Schwarz, G. J. & Kennicutt, R. C. 2004, BAAS,36, 1654

Trimble, V. 1995), PASP, 107, 977Trimble, V., Zaich, P. & Bosler, T. 2005, PASP, 117,

111

Grothkopf U. et al., Comparison of Science Metrics among Observatories


TTHEHE ESO TESO TELESCOPEELESCOPE BBIBLIOGRAPHYIBLIOGRAPHY WWEBEB

IINTERFNTERFACEACE – L– L INKINGINKING PPUBLICAUBLICATIONSTIONS

ANDAND OOBSERBSERVVAATIONSTIONS

THE ESO TELESCOPE BIBLIOGRAPHY LINKS SCIENTIFIC PAPERS BASED ON VLT OBSERVATIONS WITH UNDER-LYING OBSERVING PROPOSALS AND ARCHIVAL DATA. IT CAN BE QUERIED THROUGH A WEB INTERFACE AT ESOAS WELL AS THROUGH A FILTER AT THE NASA ASTROPHYSICS DATA SYSTEM (ADS) FROM WHICH ACTIVE LINKS

LEAD TO THE ESO ARCHIVE. THESE SERVICES ARE PREREQUISITES OF THE VIRTUAL OBSERVATORY AS THEY

ALLOW USERS TO KEEP TRACK OF THE ENTIRE LIFETIME OF SCIENTIFIC PROPOSALS, FROM SCHEDULING TO

OBSERVATIONS AND PUBLICATIONS.

NNAUSICAAAUSICAA DDELMOTTEELMOTTE 11, A, ANGELIKANGELIKA TTREUMANNREUMANN 11, U, UTTAA GGROTHKOPFROTHKOPF 11,,AALBERLBERTOTO MMICOLICOL 22, A, ALBERLBERTOTO AACCOMAZZICCOMAZZI 33

1EUROPEAN SOUTHERN OBSERVATORY, GARCHING, GERMANY2SPACE TELESCOPE EUROPEAN COORDINATING FACILITY, GARCHING, GERMANY3HARVARD-SMITHSONIAN CENTER FOR ASTROPHYSICS, CAMBRIDGE, MA, USA

TTHE ESO TELESCOPE bibliogra-phy started out in the late sev-enties as a compilation of pa-pers authored by ESO staff and visiting astronomers at the

ESO telescopes in Chile. These lists, pub-lished in the ESO Annual Report, providetoday the only record of papers based on ESOobservations arising from those early years.

In the early nineties, ESO began to storebibliographic information of relevant papersin a database which is now maintained by the library. Later, the focus shifted from bib-liographic details to observing facilities. In1996, we started to associate telescope infor-mation to all papers based on data from ESOLa Silla. Instrument information has beenadded to La Silla papers since 2002, and topapers using VLT data from First Light on-wards.

Furthermore, all VLT-based papers refer-ence the programmes from which the dataoriginated through programme identificationcodes (programme IDs), the unique iden-tifiers of ESO observing proposals. Pro-gramme IDs are the vital link between archiveand pre-observing (proposal and phase II)information and will become a key compo-nent in joining the various ESO databases inpreparation for the Virtual Observatory.

HARVESTING DATA

Today, the ESO telescope bibliography is arelational database that contains refereedpapers based on new or archived ESO data.Papers that merely cite results published inprevious papers are not included. We alsoexclude conference proceedings, even if pub-lished in refereed journals. The compilationprocess was described in detail in a recent

Messenger article (Leibundgut, Grothkopf &Treumann 2003) as well as in Grothkopf etal. (this issue, page 45).

THE WEB INTERFACE

As of July 2004, public access to the databasebecame available through a web interface,located at http://www.eso.org/libraries/telbib.html (see Figure 1). This is the resultof a collaboration between the ESO librarystaff and the DMD/DFO Database ContentManagement team within the ESO Archive.New entries in the database as well as poten-tial updates of already present records aremade available to the community via an auto-matic process that runs every night.

The general layout of the web interface is a query form that offers search options bybibliographic and observational criteria. Likemost ESO/ST-ECF Science Archive Facilityquery forms, it is based on WDB, a softwarededicated to building interfaces between theWorld Wide Web and SQL databases (Ras-mussen 1995). One of the main advantagesof WDB is the possibility to convert data fromand to the database. Data retrieved can beconverted for instance into hypertext links,which provide an easy mechanism for join-ing databases. Thus, the telescope bibliog-raphy is now linked to the ESO observingprogramme and scheduling interface of theESO/ST-ECF Science Archive Facility.

QUERY PARAMETERS

The query form is divided into three areas:Bibliographic Information, Observation In-formation and Display Options. The Biblio-graphic Information section offers searchesbased on authors, title, journal and volume aswell as by uniform bibliographic code (Bib-

Code1). In the Observation Information sec-tion, queries can be limited by specific ob-serving facilities on three hierarchical levels:site/archive, telescopes and instruments. TheProgram ID(s) field allows one to retrievepapers based on ESO programme identifica-tion codes.

For each area, various query parametersare available, either as text fields, checkbox-es or pull-down menus. A checkbox in frontof each parameter defines whether or not itwill be displayed on the result page. Bydefault, all bibliographic information as wellas the instruments and programme IDs willbe shown, while the observing sites andtelescopes will not. These settings can bechanged according to the users’ preferences.Detailed search instructions are located athttp://archive.eso.org/eso/publications.htmland can be accessed by following the hyper-linked search option labels. Further onlinehelp on how to specify queries and use oper-ators is available through the “query Help”button on the main search screen.

Result lists can be arranged in a numberof ways by using the Display Options. Theoutput format is either HTML or plain text.If more than one record is retrieved, informa-tion will be displayed in tabular format, onepaper per line (Figure 2).

The Paper ID takes users to the full re-cord display. By following that hyperlink, allavailable information will be displayed in asingle HTML frame (Figure 3), regardless ofthe parameters selected on the query page.

1 For information on BibCodes see http://cdsweb.u-strasbg.fr/simbad/refcode/section3_2.html


DIGITAL WEAVING

Unless the default display settings werechanged, the query results page will providefurther hyperlinks that point towards exter-nal databases:

ADS links to the abstract at the ADSAbstract Service (mirror located at ESO,http://esoads.eso.org/abstract_service.html).From there, the full refereed article can beretrieved (provided the library subscribes tothe journal), typically either in HTML or inPDF format. BibCodes (see above) play amajor role here in joining the data reposito-ries of ESO and ADS. Linking scattered infor-mation among multiple databases and sys-tems is part of information integration andwill ultimately enable data discovery amongheterogeneous datasets. This is possiblethanks to the development of interoperabili-ty standards and tools for which the BibCodeusage is a first example.

Program IDs provide access to observingprogramme information and to the underly-ing observations (Figure 4). As of October2004, abstracts of proposals are made avail-able once the proprietary period of the asso-ciated data has expired. In addition, links toraw and possibly reduced data as well as toother publications related to the same pro-gramme are provided.

The reverse link has also been implement-ed. Starting at the ESO observation sched-ule query form at http://archive.eso.org/wdb/wdb/eso/sched_rep_arc/form, observing pro-gramme information can be retrieved by var-ious qualifiers. From the results pages, userscan reach the list of publications associatedwith a given VLT programme ID.

ACCESS THROUGH ADSAnother way of retrieving papers using ESOtelescope data is via the ADS Abstract Ser-vice. On the main ADS search screen, userscan scroll down to the Filters section andchoose Select References In / All of the fol-lowing groups: ESO / Telescopes. In addition,at least one other search criterion must beentered, for instance a publication year or anauthor’s name.

ADS harvests new entries in our data-base at least once per week from an ESO webpage. Unlike the ESO interface, however,ADS does not cater for queries limited to spe-cific telescopes and instruments; only theentire database can be searched.

Linking datasets and publications is cur-rently a much discussed issue and proceduresare being designed for future papers to re-fer back to the datasets related to them(Accomazzi & Eichhorn 2004). In this con-text, ADS has already implemented an optionto access accompanying online data stored indata centers. This option can be activated byfollowing the letter D hyperlink (“On-lineData”) provided on ADS query result pages.For instance, if ESO data are associated witha paper, the D link takes users into the ESO

Figure 1 (above): ESOtelescope bibliographyweb interface athttp://www.eso.org/libraries/telbib.html.

Figure 2:Sample results page.

Figure 3:Full record display(excerpt).


Figure 4: ObservingProgramme queryresult page (excerpt).Entries provide accessto the proposal ab-stract, raw and possiblyreduced products aswell as further publica-tions resulting from the same proposal.

telescope bibliography from which proposaland observation information can be accessedas described above.

NEW CHALLENGE FOR THE

VIRTUAL OBSERVATORY

Online data in astronomy are increasing dra-matically. The major astronomical journalsare available in electronic format. New meth-odologies for information retrieval have to beapplied in order to exploit these data reposi-tories in the context of the Virtual Obser-vatory. Implementing interoperable archivesand communication protocols is a major taskin order to enable knowledge discovery. Web-

based technologies and new user interfacesare part of this approach for which the ESOtelescope bibliography is one example. It serves several purposes:– Bibliometrics and measuring scientific

success by deriving statistics on the num-ber of papers published per hierarchicallevel, e.g. observatory, telescope, instru-ment or programme ID.

– Access to proposal information, schedul-ing and archived raw data based on biblio-graphical references.

– Setting the pace towards the ongoing Vir-tual Observatory through the developmentof inter-connected databases.

The ESO telescope bibliography also endowsmaximum return of science benefits fromobserving proposals as it fulfils the basicrequirement of providing access to each pointof their life cycle.

REFERENCESAccomazzi, A. & Eichhorn G. 2004, ADASS XIII,

ASP Conference Series, vol. 314, 181Grothkopf, U., Leibundgut, B., Macchetto, D. et al.

2005, The Messenger, 119, 45 Leibundgut, B., Grothkopf, U. & Treumann A.

2003, The Messenger, 114, 46Rasmussen, B. F. 1995, ADASS IV, ASP Confer-

ence Series, vol. 77, 72Schmitz, M., Helou, G., Lague, C. et al. 1995, Vis-

tas in Astronomy, vol. 39, issue 2, 272

ESO EESO EXHIBITIONSXHIBITIONS ININ GGRANADARANADA ANDAND NNAPLESAPLES

EEDD JJANSSENANSSEN, EUROPEAN SOUTHERN OBSERVATORY

From September 13–17 2004, the annualJENAM (Joint European and National Astro-nomical Meeting) conference was held at the Palacio de Congresos in Granada, Spainunder the title ‘The many scales of the Uni-verse’. At the conference, ESO maintained a65 sqm exhibition stand showing the mostrecent scientific and technical achievementsat the organisation. The conference wasattended by over 450 participants, but alsomedia and local politicians visited the event.The ESO stand drew great attention bothfrom the conference participants and themedia, resulting in several articles and tele-vision broadcasts in Spain.

Given Spain’s participation in the ALMA proj-ect and its interest in joining ESO, it is hardlysurprising that among Spanish visitors andconference participants there was a stronginterest in the presentation of ALMA and theOWL project, as well as in ESO in general.Many questions were asked and the attend-ing ESO staff was busy with providing ad-ditional information material about our orga-nization.

ESO at the ‘FuturoRemoto’ exhibition.

Jorge Melnick givingone of many interviewsat the JENAM confer-ence in Granada.

From November 10, 2004 to January 30,2005, ESO took part in the ‘Futuro Remoto’exhibition, held at the Città della Scienza in Naples, Italy. Città della Scienza is the firstItalian science centre, an innovative muse-um where visitors can learn about science inan interactive way. Every year more than150 000 guests, especially school pupils,visit the museum.

‘Futuro Remoto’ is a yearly multimedia eventfor the advancement of scientific and tech-nological culture. In the past 17 years it hasstrongly contributed to bring the studentsand the citizens closer to scientific researchand technological innovations. This is testi-fied by an ever increasing flux of public, rich-ness of programmes and media coverage.

The event consisted of exhibitions coveringabout 3 000 sqm as well as a series of public lectures by well-known scientists fromItaly and abroad, including Seth Shostak,Margherita Hack, Paolo Nespoli and Massimo Capaccioli. Shows, workshops, in-teractive demonstrations, etc. completedthis successful event. With about 31 000 visi-tors in the short period between 10th and28th November, the organisers decided toextend the period of opening by a week. The success of ‘Futuro Remoto’ clearly de-monstrated the strong interest in science by the Italian public and ESO was pleased toplay a role in this public outreach activity.


TThe European Association forAstronomy Education (EAAE)held its triennial General As-sembly at ESO Headquarterson March 4–6, 2005. The for-

mal sessions were embedded in a meeting on“New Teaching Opportunities in Astronomy”which exposed more than 70 participantsfrom all over Europe to the latest didactictrends in this area. The EAAE was created in1995 and has been closely associated withESO ever since in a very fruitful collabora-tion with the ESO EPR Department and itsEducational Office.

The meeting was opened by EAAE Pres-ident, Fernand Wagner (Luxemburg), whogave an overview of current EAAE activities,together with other officials of the Associa-tion. This being the International Year ofPhysics and the centenary of Einstein’s trail-blazing research papers, much attention wasgiven to the question of how to incorporatethe related physics topics into current schoolcurricula. This is clearly not an easy task –relativity is an elusive subject for young peo-ple (and the public). Fortunately, there areseveral applications in daily life, e.g. GPSnavigation. There is no lack of connectionswith astronomical objects either and this sci-ence is thus a very useful medium (gravita-tional lenses, black holes) for discussing thissubject. Other sessions dealt with exoplanets,one of the hottest topics in present astro-physics and always of great interest to schoolstudents for obvious reasons.

EAAE and ESO are currently involved in sev-eral joint educational projects. For one ofthese, “Catch a Star! 2004”, the winners wereannounced via a Webcam; the five top prizesare trips to various observatories and went toteams from Spain, Bulgaria, France, Ger-many and Belarus.

The ALMA Interdisciplinary TeachingProject (ALMA ITP, see Messenger 117, pp 63–65) met with much interest, and thepossibility to link many different teachingtopics within one specific research project isconsidered as one of the most promising pathstowards better and more exciting scienceteaching in the schools.

Plans are now being drawn up for a Euro-pean Astronomy Day in 2006 (most proba-bly on Friday, October 20), with associat-ed opportunities for active participation ofschool students and their teachers. During theensuing discussion many useful suggestionswere made for international projects on thisoccasion.

It has become increasingly clear that inorder to raise and maintain the interest ofyoung people in science, it is necessary tobegin as early as possible, addressing the agegroup of 5–9 year-olds. This implies scienceteaching already in the primary school. A fullsession therefore dealt with this topic and theparticipants received positive reports fromseveral pilot projects. During three long ses-sions on new teaching methods, severalteaching projects at the foremost frontlinewere presented and thoroughly discussed,

some involving semi-professional equipmentlike large CCD’s and image processing, ro-botic telescopes, etc. There is no doubt thatdoing science “ (almost) like the scientists”is a great stimulus to the students.

The participants came away from themeeting with a good feeling that many newdevelopments are now taking place and thatthere is much new material and experiencewhich may advantageously be introducedinto the classes. The EAAE members lookforward to further intensifying their efforts toimprove astronomy education in Europe’sschools and ESO, through its EducationalOffice, will continue to support the EAAE inthe best possible way. Indeed, this direct“link” between the world’s top telescopes andEurope’s schools has already proved its worthin numerous ways, stimulating science teach-ing in many places.

EAAE MEAAE MEETINGEETING AATT ESO HESO HEADQUAREADQUARTERSTERS

RRICHARDICHARD WWESTEST, EUROPEAN SOUTHERN OBSERVATORY

The ‘Catch a star’ (CAS 2004) competition is enjoy-ing ever increasing popularity in schools acrossEurope. The announcement of the prize winners forCAS 2004 took place during the EAAE meeting – inan animated atmosphere as can be seen in thesepictures.


PPUBLICUBLIC AAFFFFAIRSAIRS HHIGHLIGHTSIGHLIGHTS ININ CCHILEHILE ::AACHIEVEMENTSCHIEVEMENTS OFOF 20042004

VVALENTINAALENTINA RRODRÍGUEZODRÍGUEZ ANDAND FFELIXELIX MMIRABELIRABEL ,,EUROPEAN SOUTHERN OBSERVATORY

OONE OF THE PRIORITIES OF ESOis to contribute to the aca-demic, educational and cul-tural development in Chile,fostering excellent relation-

ships with its people and government at alllevels. Here we present three 2004 highlightsof our extensive outreach programs in Chile.

PRESIDENT LAGOS VISITS PARANAL

The ESO Paranal Observatory was honoredwith a very special overnight visit on Decem-ber 9th to 10th, 2004. The President of Chile,Ricardo Lagos Escobar and his wife LuisaDurán, arrived to the VLT for the first timeand were hosted by the ESO Director Gener-al, Dr. Catherine Cesarsky; ESO’s Represen-tative in Chile, Mr. Daniel Hofstadt; Prof.Maria Teresa Ruiz, Head of the AstronomyDepartment at the Universidad de Chile, andESO staff members. The distinguished visi-tors were shown the various high-tech instal-lations at the observatory, including the Inter-ferometric Tunnel with the VLTI delay linesand the first Auxiliary Telescope. Explana-tions were given by ESO astronomers andengineers and the President, a keen amateurastronomer, gained a good impression of thewide range of exciting research programs thatare carried out with the VLT. President Lagosparticipated in the imaging of the barred spi-ral galaxy NGC 1097, located at a distance ofabout 45 million light-years (see cover).

INTERACTIVE ASTRONOMY

One of the most popular stands at the Inter-active Museum in Santiago (MIM) is the newAstronomy Exhibition developed by ESO incollaboration with MIM. Since the openingon November 12th 2004, this astronomicaljourney has captured the attention of childrenand adults with four dynamic videos and aunique space walk through galaxies, nebulaeand planets. In 20 minutes, visitors not onlylearn basic concepts of astronomy such asgravitation, spectroscopy and space-time, butalso enjoy some of the most stunning astro-nomical images collected by ESO telescopesin Chile.

PLANETARY ECOLOGY SUCCESS

Since 1997 ESO has sponsored in Chile the PROED programs of Planetary Ecology,Helios and Universum, making astronomymore accessible for students and teachers ofpublic schools from the II, III and IV Regionsin Chile.

ESO support has made it possible to trainteachers, develop kits of didactic materials,provide access to amateur telescopes, astron-omy exhibitions, regional and national Work-shops on Astronomy for schools and theestablishment of science-teachers networks.

The success of these programs is based oninnovative and participative methods, includ-ing experiments in the classroom with the stu-dents.

An example of this is the Planetary Ecol-ogy program, almost completely financed byESO for children between 8 and 12 years.There are already 268 trained teachers in 153schools using this program to teach studentsabout space, earth, water, air and the ecosys-tem in an imaginative way.

President RicardoLagos, ESO DirectorGeneral, Dr. CatherineCesarsky, and Dr. Jason Spyromilio,future Director of La Silla-Paranal obser-vatories.

ESO’s Representativein Chile, Daniel Hof-stadt, and MIM Direc-tor, Haydée Domic, join Mrs. Luisa Duránde Lagos (in the centre)during the opening of the Astronomy Ex-hibition.

Chilean students of the Planetary Ecologyprogram supported by ESO.


like being in the jungle! So much humidity,compared to outside!

My scientific work is the understanding ofour Galaxy’s childhood. There are mainlytwo ways to do so. Either you look at verydistant extragalactic objects, or you look atvery old galactic ones. That’s what I do.

Some low mass stars have atmospheresthat are not evolving with time. Thus, bystudying the chemical composition of theseatmospheres, we have a quite direct access tothe material from which these stars wereformed. As we suppose that these stars wereformed very early after the Big Bang, thechemical composition of these atmospheresputs strong constraints on the Big Bang andthe supernovae models. And a striking thingconcerning these atmospheres is that at thattime, we have been able to detect half of theelements of the periodic table – elementsfrom Hydrogen to Uranium!

MARKUS HARTUNG

IT’S FASCINATING TO

sit in the controlroom, piloting aworld-class cutting-edge telescope! Asingle mouse clickand 400 tonnes aremoved to point to the

new target. Between the active optics givingthe 8-m-primary-dish its optimum shape anda deformable mirror countersteering 500times per second to correct atmospheric tur-bulence in real time we finally recover space-like resolution. And the more you look intothe details, the more you are amazed that allthese systems work reliably together to com-prise one of the sharpest eyes with which toexplore the universe.

This is about the thrill of being a fellow atthe VLT. You probably won’t believe thatbeing a fellow is all fun and games. Well, Iadmit that the work can be very demandingand stressful, and after a long shift there canbe friction. Paranal science operations seemsultra-organized – rules and regulations every-where. One might forget that they are notthere to bother but to help people work togeth-er successfully – no doubt about their posi-tive pay back! When I take care of visitingastronomers as part of my fellow's duties, itis a pleasure to see most of them leavinghappy and impressed by the observatory. Theteam spirit is great, and I really enjoy the priv-ilege of working with such ambitious peoplefrom all over the world.

I started as an ESO Paranal fellow with myduty station in Santiago in June 2003. Before,I had been working on my PhD at the MPIAin Heidelberg and had strongly contributed to

FFELLOWSELLOWS AATT ESOESONICOLAS BOUCHÉ

I ARRIVED AT ESOGarching in Septem-ber 2003 as a post-doc under a jointMPA/ESO contractfunded by the Eu-ropean CommunityResearch Training

Network (“The physics of the intergalacticmedium”) just after graduating from the Uni-versity of Massachusetts. Since I was little –when Halley’s comet last came around to beprecise – I wanted to become an astronomer.Given that I grew up in Belgium I knew aboutESO and its state of the art facilities. It wasthus a great privilege to start my career at ESOheadquarters. Of course my interests as a pro-fessional astronomer took me as far as youcan from comets. I now study galaxy forma-tion and distant galaxies at redshifts of two tothree. More specifically, my thesis was cen-tered on using cross-correlation techniques tomeasure (and put constraints on) the mass ofa class of high-redshift gas clouds, the so-called Damped Lyman-alpha Absorbers.

I was very fortunate that during my stay atESO, many fellows worked in my field,namely, the field of quasar absorption linesystems. This vibrant environment led to newcollaborations, many, many helpful discus-sions and new friendships. I also benefitedgreatly from the institutions around ESO, theMax-Planck Institut für Astrophysik and theMax-Planck Institut für extraterrestrischePhysik where I just started a three year con-tract. These three institutions are at the fore-front of European astronomy, are leadingmany cutting edge experiments and it’s agreat pleasure to be part of it.

Even though, when the fog rolls in for twomonths at a time, it’s like living in a subma-rine, Bavaria is a beautiful region with manypossibilities centred on the Alps or just inMunich.

ERIC DEPAGNE

I AM AT THE MOMENT

right in the middle of my four yearscontract as a fellow in Chile. I have myduties at Paranalwhere, as a night as-tronomer, I do the

observations on UT2 (FLAMES, UVES andFORS1).

Paranal is an incredible place to see. Aftera two hour drive from Antofagasta, in themiddle of the desert, you can see the four tele-scopes on top of the beheaded mountain.Then, you enter the residencia. And you feel

the CONICA project. The marriage of thisversatile infrared camera with the NAOSadaptive optics system took place in Paris. Ispent one year there integrating the instru-ments before they were shipped to Paranaland commissioned at the fourth VLT tele-scope. When I joined ESO, I had alreadyworked for several months at Paranal. Then,I did not get a chance to explore Chile, but Ihoped in my heart that I would return to catchup with it. And here I am now – with my wifeand my (Chilean) son, born last December inSantiago – it’s been a truly exciting experi-ence!

JOCHEN LISKE

FOR MY PHD I decid-ed to be adventurousand exchanged thehomely shores of theriver Rhine (Bonn)with the sandybeaches of the SouthPacific. At the Uni-

versity of New South Wales in Sydney, Aus-tralia, I worked on my beach volleyball skillsas well as the intergalactic gaseous structurescalled “quasar absorbers”, which provide awealth of information on the physicalconditions in the early Universe, studyinglarge-scale structure, the radiative environ-ment of powerful quasars and weighing the“normal” matter content of the Universe.

At the end of my PhD in 2001 I moved tothe misty shores of the Firth of Forth (Edin-burgh) to explore a different area of researchand to begin work on the Millennium GalaxyCatalogue, a large-scale survey of localgalaxies. When trying to understand howgalaxies evolve with time we need to com-pare today’s galaxies with those in the earlyUniverse. It is particularly important to beable to distinguish the various constituents ofgalaxies as they are believed to be the resultof different formation processes. The Millen-nium survey provides a detailed picture of(cosmologically speaking) nearby galaxiesand hence forms the “today” part of the abovecomparison.

Last year I moved to the banks of theMühlbach (ESO) where I am continuing mywork both on quasar absorbers and galaxyevolution. I am also supporting SINFONI, anew VLT instrument currently being com-missioned. Recently, I got involved with aworking group who are considering the useof OWL for an experiment designed to liter-ally watch the Universe change its expansionspeed over the timescale of a decade or so!Crazy? Possibly, but great fun! This is the sortof thing that makes ESO a fascinating place.Here you get to see the wheels of astronomyturn and watch the future take shape.


TTHEHE UKIRUKIRT IT INFRAREDNFRARED DDEEPEEP SSKYKY SSURURVEYVEY (UKIDSS): (UKIDSS): DADATTAA ACCESSACCESS

AANDYNDY LLAAWRENCEWRENCE, UNIVERSITY OF EDINBURGH

SSTEVETEVE WWARRENARREN, IMPERIAL COLLEGE LONDON

As part of the UK accession to ESO, it wasagreed that all astronomers in ESO memberstates would have access to the data from the UKIRT Infrared Deep Sky Survey(UKIDSS – http://www.ukidss.org). This ex-citing survey is 3 magnitudes deeper than2MASS over 7 500 square degrees, as well ashaving small deep areas, all the way down toa one square degree Ultra Deep Survey to K = 23. UKIDSS is therefore the real equiv-alent to optical sky surveys such as those from the Palomar, UK and ESO Schmidt tele-scopes, and the SDSS (see The Messenger,108, 31). A substantial fraction of the surveyis in the southern sky. The survey is being car-ried out by the new UKIRT Wide Field Cam-era (WFCAM), which has just been commis-sioned – see the press release at http://outreach.jach.hawaii.edu/pressroom/2004-wfcam/). The survey starts in March and willtake seven years. There will be a sequence ofreleases, each of which is available only toESO astronomers (and a small additional listof individuals in Japan) for eighteen months,

following which the data become publicworldwide.

Access to reduced data and survey prod-ucts will be through a queryable sciencearchive known as the WFCAM ScienceArchive (WSA: http://surveys.roe.ac.uk/wsa). The WSA link is not live until the firstrelease, but you can try out the interface byusing the SuperCOSMOS Science Archive(SSA: http://surveys.roe.ac.uk/ssa). Duringthe ESO-proprietary period, you will need toregister for a username and password toobtain access. However the WSAteam do notexpect to vet and authorise every astronomerin Europe who potentially might want accessto UKIDSS, and furthermore this list willchange from month to month. Rather, we willbe agreeing a list of “community contacts” towhom we will delegate the task of producinga list of authorised local usernames. Ingest-ing these into the WSA database will be auto-mated, so that updating these lists as peopleleave and join should be easy to do at fairlyregular intervals. “Community” will normal-

ly mean one particular university or observa-tory, but could potentially mean any welldefined group of astronomers.

We have set up a preliminary list of com-munity contacts which can be found on theUKIDSS web site at http://www.ukidss.org/archive.html. If you think your organisationor department needs to be added, please talkamongst yourselves to choose a contact, andsend an email to the UKIDSS registrationteam below. Note that this can be done at anytime from now on – the list can grow slowly.As soon as we have an agreed initial list, wewill send out instructions for registering yourmembers.

UKIDSS Registration Team:Andy Lawrence, Edinburgh, [email protected]

(UKIDSS Consortium PI) Steve Warren, Imperial College London,

[email protected] (UKIDS Con-sortium Survey Scientist)

Nigel Hambly, [email protected] (WSA Scien-tist)

Andy Adamson, [email protected] (Head of UKIRT)

A central region of theOrion Nebula capturedin the infrared byUKIRT’s Wide FieldCamera (WFCAM),courtesy of the JointAstronomy Centre.(Data processing wasby Chris Davis andWatson Varricatt). In asingle exposureWFCAM can image a region the same area as the Full Moon.

Ann

ounc

emen

ts


SYMPOSIUM ON RELASYMPOSIUM ON RELATIVITYTIVITY, MA, MATTER TTER AND COSMOLOGYAND COSMOLOGY

11–14 July 2005, Bern, Switzerland

The EPS-ESA-ESO-CERN Symposium on ‘Relativity, Matter and Cosmology’, will take place from July 11 to 14, 2005 at the University of Bern.The Symposium will be part of the centenary celebrations of Albert Einstein’s annus mirabilis and take place in the framework of the 13th General Conference of the European Physical Society, EPS13, under the general title ‘Beyond Einstein – Physics for the 21st Century’. TheEPS-ESA-ESO-CERN Symposium 2005 is expected to become one of the highlights of the ‘World Year of Physics 2005’, also declared the‘International Year of Physics’ by the United Nations.

Plenary speakers will introduce the topics, and ample time and space will be provided for contributed papers and posters describing originalwork, preferably by young scientists.

Plenary Speakers in the EPS-ESA-ESO-CERN Symposium on ‘Relativity, Matter and Cosmology’ will include K. Danzmann, G. Drexlin, G. Efstathiou, J. Engelen, C. W. F. Everitt, E. Fiorini, W. Gelletley, F. Iachello, V. M. Kaspi, G. Ross, B. F. Schutz, J. Silk, D. Spergel, J. Stachel,and F. Wagner.

The following Sessions for contributed papers and posters are foreseen:

– The Fundamental Laws of Physics and the Constancy of Fundamental Constants

– Tests of Gravitational Theory and General Relativity– Quantum Gravity– Dark Energy– Gravitational Waves– String Theory and Extra Dimensions– The Standard Model and Beyond

EPS13 comprises two other co-located Conferences, one on ‘Photons, Lasers and Quantum Statistics’ and another one on ‘Brownian Motion,Complex Systems and Physics in Biology’. It is anticipated that each one of these Conferences will attract about 300 physicists and astronomers,not only from Europe, but also from other continents. Registration for EPS13 will give access to all three conferences, which will be run in par-allel and in close proximity to each other.

Further information can be found at http://www.eps13.org

– LHC Physics and the Origin of Mass– Neutrino Oscillations and Masses– Matter in Extreme Conditions– Dark Matter– The Early Universe– Cosmological Parameters– Matter in the Universe– Supernovae in Cosmology

ESO WESO Workshop onorkshop on

MMULULTIPLETIPLE SSTTARSARS ACROSSACROSS THETHE H-R DH-R DIAGRAMIAGRAM12–15 July 2005, ESO Headquarters, Garching, Germany

Multiple (i.e. triple, and higher order) stellar systems comprise a significant fraction of stellar populations. Their role in stellar physics has notyet been fully recognized. Stars of some specific types can originate only in multiple systems. An interplay between tides, nuclear evolutionand dynamics is a rich field of contemporary research. The growing observational data on multiple systems with a variety of characteristics isused to critically examine the assumptions underlying stellar evolutionary models.

The main aim of the workshop is to bring together observers using different techniques (e.g. spectroscopy, high angular resolution imaging),from X-rays to far-IR, on ground-based single telescopes or interferometers, and on space observatories. The combination of techniques isvital for comprehensive studies of multiple stars that span a wide range of angular separations and stellar types.

The current state of observational and theoretical knowledge will be reviewed. Priorities for future studies will be identified, so as to providethe necessary input for further progress in the understanding of the genesis of multiple stars, their structure and their role for the study of stel-lar evolution.

The format of the meeting will consist of invited talks, contributed talks and posters.

SOC members: J. Bouvier, L. M. Close, P. P. Eggleton, R. F. Griffin, W. I. Hartkopf, S. Hubrig (co-chair), Ch. Leinert, M. Petr-Gotzens, M. Sterzik,A. Tokovinin (co-chair), S. Udry

LOC members: Ch. Stoffer, P. Bristow, A. Kellerer, M. Petr-Gotzens

Full details and registration information can be retrieved from http://www.eso.org/gen-fac/meetings/ms2005/ or by e-mail to [email protected]

Deadline for registration: 30 April 2005


MPMPA/ESO/MPE/USM Joint AstrA/ESO/MPE/USM Joint Astronomy Conferonomy Conference on ence on

OOPENPEN QQUESTIONSUESTIONS ININ CCOSMOLOGYOSMOLOGY::THETHE FF IRSTIRST BBILLIONILLION YYEARSEARS

22–26 August 2005, Garching, Germany

The accepted lore about the state of modern cosmology is that existing data constrain models well enough that few important questions remainto be either asked or solved. In the last year, a wealth of new data about the cosmological background radiation, early star and galaxy forma-tion as well as cosmological evolution in general has been generated. The theoretical prejudice is that the current concordance cosmologicalmodel provides a description of our universe consistent with most of these observations. But does it provide the definitive framework againstwhich these new observations can be tested? This meeting will present the latest data, confront them with conventional and alternative mod-els, and review emerging technologies and expected data for their potential to challenge our current understanding of the universe.

Invited Speakers: K. Adelberger (OCIW), J. Bergeron (IAP), V. Bromm (U. Texas, Austin), P. Cox (IAS), L. Cowie (U. Hawaii), R. Ellis (Caltech),X. Fan (U. Arizona), A. Ferrara (SISSA), K. Gorski (JPL), Z. Haiman (Columbia), P. Jakobsen (ESA), K. Jedamzik (U. Montpellier), C. Lawrence (JPL), P. Madau (UCSC), S. Malhotra (STScI), K. Olive (U. Minnesota), J.-L. Puget (IAS), S. Ryan (Open University, UK), A. Songaila (U. Hawaii), J. Schaye (IAS), R. Schneider (OAA), C. Steidel (Caltech), F. Walter (MPIA), S. D. M. White (MPA)

Scientific Advisory Committee: X. Fan (U. Arizona), A. Ferrara (SISSA), K. Gorski (JPL), P. Jakobsen (ESA), B. Leibundgut (ESO), A. Loeb (Harvard), J. Silk (Oxford), A. Songaila (U. Hawaii), M. Umemura (U. Tsukuba), S. D. M. White (MPA)

Local Organising Committee: A. J. Banday, B. Ciardi, S. D. M. White (MPA); W. Freudling, B. Leibundgut, P. Shaver (ESO); M. Lehnert (MPE);E. D’Onghia (USM)

Further information can be found at: http://www.mpa-garching.mpg.de/~cosmo/2005/

Deadline for registration and abstract submission: 20 May 2005

ESO ConferESO Conference on ence on

SSCIENCECIENCE PPERSPECTIVESERSPECTIVES FFOROR 3D S3D SPECTROSCOPYPECTROSCOPY10–14 October 2005, ESO Headquarters, Garching

The expansion in 3D instrumentation – obtaining spatially resolved spectra over an area of the sky – on large telescopes has been phenome-nal over the past few years. Examples at ESO on the VLT are SINFONI, the Integral Field Unit (IFU) of VIMOS, the Argus mode of FLAMES; atGemini the GMOS instrument has an IFU mode; OSIRIS is soon to reach Keck. There are also powerful consortium instruments such asSAURON and CIRPASS which were designed for rather specific problems. Several new tunable filters will see first light this year and sterlingwork continues to be carried out by facility IFU instruments on 4 m class telescopes such as at La Palma (INTEGRAL) and Calar Alto (PMAS).Accompanying this leap in instrumental capabilities, the volume of data delivered is set to increase as larger or multiple IFU’s are planned,such as the MUSE or KMOS instruments for the VLT.

It is over five years since the last conference on 3D spectroscopy and time to assess where the technique is going and what are the currentand future perspectives for science from 3D spectrographs. This conference will focus on the current science and the plans for the near future.All areas of astrophysics will be covered for which 3D spectroscopy brings special benefits – from the Solar System out to high redshift galax-ies. There will be brief presentations of new instruments and techniques but the primary emphasis will be on the science from 3D spectroscopy.

The conference is co-sponsored by Euro3D, which is a European Union funded Research Training Network. Euro3D currently employs 10 post-docs at centres throughout Europe to work on integral field spectroscopy (http://www.aip.de/Euro3D/).

The meeting will consist of longer invited reviews with shorter contributed talks and posters. The proceedings will be published in the SpringerESO Astrophysics Symposia series.

SOC members: J. Allington-Smith (Durham), S. Arribas (IAC, STScI), A. Bunker (Exeter), R. Bacon (Lyon), G. Cecil (North Carolina), R. Davies (Oxford), C. Dumas (ESO), P. Ferruit (Lyon), R. Genzel (MPE), J. Bland-Hawthorn (AAO), A. Quirrenbach (Leiden), G. Wright (Edinburgh)

LOC members: J. Walsh (ESO), M. M. Roth (AIP), M. Kissler-Patig and C. Stoffer (ESO)

Further details are available at http://www.eso.org/3Dspec05 or by e-mail to [email protected]

Deadline for pre-registration: 15 May 2005Deadline for final registration: 01 September 2005


FFELIXELIX MMIRABELIRABEL BECOMESBECOMES ESO RESO REPRESENTEPRESENTAATIVETIVE ININ CCHILEHILE

Starting on April 1, Felix Mirabel will takeup a double role as ESO Representative inChile and Head of ESO’s Office for Sciencein Chile.

Felix Mirabel was born in 1944 in Uru-guay. After finishing secondary school, hewent to Argentina where he obtained a PhDin Astrophysics at the University of La Plata,and a Master degree in Philosophy from theUniversity of Buenos Aires. He pursued post-doctoral studies at Jodrell Bank in the UK, atthe University of Maryland, and then in Puer-to Rico using the Arecibo radiotelescope. In1990, after having worked at Caltech as aGuggenheim fellow, he moved to Franceworking as director of research at the CEA(Saclay), where he stayed until joining ESO.

His scientific career has spanned the elec-tromagnetic spectrum, from his early work inthe radio domain to the optical, then infrared,millimetre and X- and gamma-ray astrono-my. He has made use of a wide variety of theworld’s observational facilities, including allof ESO’s major telescopes, from the 3.6 m tothe SEST, from the NTT to the VLT.

His research has covered topics such asstar formation, galactic structure, compactbinaries, gamma-ray bursts and interactinggalaxies. He is particularly well-known for

the identification of ultra-luminous infraredgalaxies as a new class of objects and for hisdiscoveries of microquasars in our galaxy, forwhich he has been awarded prizes in the USand in Europe, and last year a Doctorate Hon-oris Causa from the University of Barcelona.

On April 1, Felix Mirabel will take up adouble role: he will be ESO’s Representativein Chile as well as the Head of ESO’s Officefor Science in Chile. As ESO Representative,his main wish is to continue to improve thepresent very good relationship between ESOand Chile, both with its people and authori-ties. He feels that “this is a very good time totake up this post, following the excellentwork done by Daniel Hofstadt.” He sees con-tinued strengthening of the relations withChile not only through an enhancement ofscientific capabilities of Chilean universitiesbut also through cultural and educational pro-grammes at large. In his post, Felix Mirabelwill also strive to work in harmony with theUS and Japan for the construction of ALMA.

In his other role as Head of ESO’s Officefor Science in Chile, his aim is to fosterscience across all ESO centres in coopera-tion with Chilean research groups. “When I arrived, I was impressed by the work of Danielle Alloin. Together with Daniel

Hofstadt, she has shaped Vitacura and madeit an attractive astronomical centre in Chile.I also appreciate very much the fact that allESO staff members, fellows and students areeager to enhance science in Chile, organizingby own initiative thematic groups of discus-sion and helping in a great variety of activi-ties. Certainly, my new tasks won’t leave memuch time to continue my own research inthe same way as before. But through this newrole in the organization I aim to work forexcellent conditions of astronomical researchin Chile, and I am convinced that I will great-ly enjoy the discoveries to be made at ESOby the young generation of astronomers.”

ESO SESO STUDENTSHIPTUDENTSHIP PPROGRAMMEROGRAMME

The European Southern Observatory research student programme aims at providing opportunities to enhance the Ph.D. programmes of ESOmember-state universities. Its goal is to bring young scientists into close contact with the instruments, activities, and people at one of theworld’s foremost observatories. For more information about ESO’s astronomical research activities please consult http://www.eso.org/science/The ESO studentship programme is shared between the ESO headquarters in Garching (Germany) and the ESO offices in Santiago (Chile).These positions are open to students enrolled in a Ph.D. programme at a university in an ESO member state or, exceptionally, at an institutionoutside ESO member states.Students in the programme work on their doctoral project under the formal supervision of their home university. They come to either Garchingor Santiago for a stay of normally one or two years to conduct part of their studies under the co-supervision of an ESO staff astronomer. Can-didates and their home-institute supervisors should agree on a research project together with the ESO local supervisor. A list of potential ESOsupervisors and their research interests can be found at http://www.eso.org/science/sci-pers.html#faculty and http://www.sc.eso.org/santiago/science/person.html. A list of current PhD projects offered by ESO staff is available at http://www.eso.org/science/thesis-topics/. Itis highly recommended that the applicants start their Ph.D. studies at their home institute before continuing their Ph.D. work and developingobservational expertise at ESO.In addition, the students in Chile have the opportunity to volunteer for as many as 40 days/night work per year at the La Silla-Paranal obser-vatory. These duties are decided on a trimester by trimester basis, aiming at giving the student insight into the observatory operations, andshall not interfere with the research project of the student in Santiago.

The closing date for applications is June 15, 2005.Please apply by: filling the form available at http://www.eso.org/gen-fac/adm/pers/forms/student05-form.pdf and attaching to your application:– a Curriculum Vitae (incl. a list of publications, if any), with a copy of the transcript of university certificate(s)/diploma(s).– a summary of the master thesis project (if applicable) and ongoing projects indicating the title and the supervisor (maximum half a page),

as well as an outline of the Ph.D. project highlighting the advantages of coming to ESO (recommended 1 page, max. 2). – two letters of reference, one from the home institute supervisor/advisor and one from the ESO local supervisor, – and a letter from the home institution that i) guarantees the financial support for the remaining Ph.D. period after the termination of the ESO

studentship, ii) indicates whether the requirements to obtain the Ph.D. degree at the home institute are already fulfilled.All documents should be typed and in English (but no translation is required for the certificates and diplomas).The application material has to be addressed to:

European Southern Observatory Studentship Programme Karl-Schwarzschild-Str.2

85748 Garching bei München (Germany)

All material, including the recommendation letters, must reach ESO by the deadline (June 15); applications arriving after the deadline orincomplete applications will not be considered!Candidates will be notified of the results of the selection process in July 2005. Studentships typically begin between August and Decemberof the year in which they are awarded. In well justified cases starting dates in the year following the application can be negotiated.For further information contact Christina Stoffer ([email protected]).


Applications are invited for the position of

HHEADEAD OFOF AADMINISTRADMINISTRATIONTION

CCAREERAREER PPAATHTH: VII: VII

Purpose and scope of the position: The main task is to provide efficient administrative services and advice to the Director General, DivisionLeaders and to staff members in the scientific and technical areas in the fields of financial planning and accounting, personnel management,purchasing, legal and contractual matters, information systems and building and site maintenance. As a member of the ESO Management theHead of Administration contributes essentially to the development of the overall policy, strategic planning, relations to the members of thepersonnel and maintains professional contacts at highest level outside the Organisation. ESO employs in total approximately 650 staff mem-bers and the Administration Division comprises the Administration at the Headquarters in Garching near Munich and the Administration in Santiago (Chile). The successful candidate will be supported by some 50 qualified staff members.Professional requirements/qualifications: An appropriate professional qualification as well as substantial management and leadership expe-rience within a scientific organisation, preferably international, and knowledge of administrative, legal, financial and personnel procedures arerequired. Excellent communication skills and a very good knowledge of English are essential. Good knowledge of the German language wouldbe an important asset. Additional knowledge of other European languages, in particular Spanish, would be an advantage.Remuneration and contract: We offer an attractive remuneration package including a competitive salary (tax-free), comprehensive pensionscheme and medical, educational and other social benefits as well as financial support in relocating your family. The initial contract is for aperiod of three years with the possibility of a fixed-term extension or permanence. Serious consideration will be given to outstanding candi-dates willing to be seconded to ESO on extended leaves from their home institutions. Either the title or the grade may be subject to changeaccording to qualifications and the number of years of experience.Staff category: International Staff Member.Duty station/Place of residence: Garching near Munich, Germany, with regular duty travels to Chile.Starting date: 1 September 2005.Applications: If you are interested in working in a stimulating international research environment and in areas of frontline science and tech-nology, please send us your CV in English and the ESO Application Form (to be obtained from the ESO Home Page at http://www.eso.org/gen-fac/adm/pers/forms/) by

30 April 2005.

For further information please contact Mr. Roland Block at [email protected] are also strongly encouraged to consult the ESO Home Page (http://www.eso.org) for additional information.

Although preference will be given to nationals of the Member States of ESO: Belgium, Denmark, Finland, France, Germany, Italy, The Nether-lands, Portugal, Sweden, Switzerland, and United Kingdom, no nationality is à priori excluded. The post is equally open to suitably qualifiedmale and female applicants.

Applications are invited for an Operations Staff Astronomer position in the Science Operations Department at the Very Large Telescope onCerro Paranal near Antofagasta, Chile. This post is open to suitably qualified men and women:

OOPERAPERATIONSTIONS SSTTAFFAFF AASTRONOMERSTRONOMER

CCAREERAREER PPAATHTH: V: V

Assignment: The successful candidate will support observing operations in both visitor and service mode at the VLT Unit Telescopes (UT) onParanal. The tasks to be performed include the short-term (flexible) scheduling of queue observations, the calibration and monitoring of theinstruments, and the assessment of the scientific quality of the astronomical data. Paranal Operations Staff Astronomers contribute to thechallenge of operating a world leading astronomical facility so as to optimize its scientific output, have the opportunity to acquire expert know-ledge of novel instrumentation, and may be given the overall responsibility for an instrument. Flexibility exists so as to tailor duties and respon-sibilities as a function of personal expertise and interests.Operations Astronomers may be members of the ESO Science Faculty, with an appointment at the level of Assistant or Associate Astronomer.They will be expected and encouraged to actively conduct astronomical research up to 50% of the time. 105 nights per year are spent at theobservatory carrying out functional duties, usually in a shift of 8 days on Paranal, 6 days off. The rest of the time is spent in the Santiago office.Depending on qualification, expertise, and personal interest, Operations Astronomers may alternatively be offered an appointment with up to20% of the time for personal research and 135 nights per year to be spent on the observatory. Financial support for scientific trips and staysat other institutions, including in Europe, is foreseen for all Paranal Operations Astronomers.Education: Ph.D. in Astronomy, Physics or equivalent.Experience and knowledge: The Observatory is seeking a staff astronomer with substantial observing experience (at least three years). Theideal candidate will be active researcher and have excellent observation oriented research records, will be familiar with a broad range of instru-mental, data analysis, archiving and observational techniques, and must be conversant with at least one major data reduction package suchas MIDAS, iraf or IDL. Of special value would be a record of instrumental experience, such as the participation in the design, construction orcalibration of existing instruments. Excellent communication skills, a good command of the English language, a working knowledge of Span-ish or a willingness to learn and a strong sense of team spirit are essential.Duty station: Paranal and Santiago, Chile. Starting date: As soon as possible. Contract: The initial contract is for a period of three years with the possibility of a fixed-term extension or permanence. Promotions will bebased on scientific as well as functional achievements. Remuneration: We offer an attractive remuneration package including a competitive salary (tax-free), comprehensive pension scheme, medical, educational and other social benefits as well as professional training opportunities and financial support in relocating your family.Either the title or the grade may be subject to change according to education and the number of years of experience.Applications consisting of your CV (in English language), the ESO Application Form (http://www.eso.org/gen-fac/adm/pers/forms/) and fourletters of reference should be submitted by

30 April 2005.

For further information, please consult the ESO Home Page (http://www.eso.org) or contact Mrs Nathalie Kastelyn, Personnel Department, Tel. +49-89-3200-6217.

Although preference will be given to nationals of the Member States of ESO: Belgium, Denmark, Finland, France, Germany, Italy, The Nether-lands, Portugal, Sweden, Switzerland, and United Kingdom, no nationality is à priori excluded.


Applications are invited for the position of an

ALMA BALMA BACKACK--ENDEND IINTEGRANTEGRATIONTION ANDAND DDAATTAA CCOMMUNICAOMMUNICATIONTION EENGINEERNGINEER


Purpose and scope of the position: Within the European ALMA team, the selected candidate will take up responsibilities in the monitoringof production contracts and in the integration, commissioning and acceptance of the Back-end sub-system and related data communicationinfrastructures (fibre optic based system) of the ALMA radio telescope. As such, the selected candidate will be part of the ALMA Back-endIntegrated Product Team (BE IPT). The position will involve duties in ALMA partner regions, Europe, North America and Chile in the form offrequent and occasionally extended missions.Duties and responsibilities: The position will include several of the following tasks:

– Elaboration of BE integration & test plans and support for the development of an operations and maintenance plan in close collaborationwith the System Engineering & Integration IPT.

– Definition of applicable Quality Management processes, mainly in the areas of production, integration and validation in the ALMA project.

– Participation in and organisation of product reviews. – Monitoring of the installation, commissioning and acceptance of the BE sub-systems and fibre optic distribution infrastructures on site.– Support for developing and maintaining the BE sub-system specifications and interfaces and for assuring that the sub-system designs

comply with defined specifications.– Monitoring of detailed sub-system performance of the BE sub-system technical budgets.– Contribution to the ALMA configuration and change-control processes.– Direct reporting to the ESO ALMA Back-end manager. Support the IPT manager in monitoring schedules and the technical performance

of all back-end subsystems. Professional requirements/qualifications: University degree (or equivalent) in Electrical Engineering or Physical Sciences. Further qualifica-tions and experience in a scientific domain would be advantageous, e.g. a post-graduate experience in radio-astronomy research and/or inhigh speed fiber optic networks and advanced electronics industrial environment. At least three years of working experience in an engineer-ing position on multidisciplinary, high technology, advanced electronics and fibre-optic data communication project(s). Experience and knowledge: The ideal candidates will have experience in:

– Design, development, production installation and commissioning of advanced electronic systems, high speed optical data communica-tion systems, optical components and related infrastructures.

– Establishing Assembly, Integration and Verification Plans communication and electronic systems. – Configuration and Quality management. – Excellent communication skills, a good command of the English language and a strong sense of team spirit are essential.

Duty station: Garching near Munich, Germany.Starting date: As soon as possible.Remuneration and contract: We offer an attractive remuneration package including a competitive salary (tax free), comprehensive pensionscheme and medical, educational and other social benefits as well as financial help in relocating your family. The initial contract is for a peri-od of three years with the possibility of a fixed-term extension or permanence. The title or grade may be subject to change according to qual-ification and the number of years of experience. Applications: If you are interested in working in a stimulating international research environment and in areas of frontline science and tech-nology, please send us your CV (in English) and the ESO Application Form (http://www.eso.org/gen-fac/adm/pers/forms/) by

15 April 2005.

For further information please contact Mrs Nathalie Kastelyn, Personnel Department, Tel. +49-89-3200-6217. You are also strongly encour-aged to consult the ESO Home Page (http://www.eso.org/) for additional information about ESO.


Applications are invited for the position of a

SSENIORENIOR EENGINEERNGINEER – H– HEADEAD OFOF PPARANALARANAL EENGINEERINGNGINEERING DDEPEPARARTMENTTMENT


Purpose and scope of the position: The role of the Paranal Engineering Department is to carry out the assembly, integration and troubleshoot-ing of the VLT (Very Large Telescope), the instrumentation of the VLT and of the VLTI (VLT Interferometry), of the VST (VLT Survey Telescope),VISTA (Visible and Infrared Survey Telescope), of all the facilities required on the Observatory (Power Station, Air Compressors, Chillers, etc),and to provide general engineering support of maintenance, troubleshooting and fault repair to nightly operations.The Engineering Department of the Paranal Observatory comprises 5 engineering groups (Mechanical, Electronic, Software, Optics, and Instru-mentation), with a total workforce presently consisting of 45 engineers and 15 technicians.The telescopes and instruments are large, complex systems that involve many advanced technologies and that require a high level of engi-neering support.Duties and responsibilities: Reporting to the Head of Engineering of the La Silla-Paranal Observatory, the successful candidate shall havethe responsibility of:

– The four telescopes with their instruments making them available for Science operations 365 nights per year;– The integration activity of new instruments and telescopes ensuring they are properly performed on time;– The engineering support to the commissioning of the interferometric complex;– Additional technical responsibilities including that of the Power Station and of the auxiliary systems (compressed air, chillers, etc.);– The definition of operational procedures and long-term planning;– The personnel management (in particular training and performance evaluation) and supervision of the five engineering groups;– The preparation of the annual budgets and the scheduling of support staff.

Professional requirements/qualifications:– University Degree in mechanical, optomechanical or aeronautical engineering.– At least ten years of relevant on field engineering experience in large multi-discipline scientific/of high technology projects;– Technical experience in optomechanic systems will be a plus.

see next page


Applications are invited for the position of an

ALMA PALMA PROJECTROJECT PPLANNERLANNER/S/SCHEDULERCHEDULER


The Atacama Large Millimeter Array (ALMA) is an international astronomy facility. ALMA is an equal partnership between Europe and NorthAmerica, in cooperation with the Republic of Chile, and is funded in North America by the U.S. National Science Foundation (NSF) in cooper-ation with the National Research Council of Canada (NRC), and in Europe by the European Southern Observatory (ESO) and Spain. ALMA con-struction and operations are led on behalf of North America by the National Radio Astronomy Observatory (NRAO), which is managed by Asso-ciated Universities, Inc. (AUI), and on behalf of Europe by ESO.The ALMA construction project has adopted a management structure based on the Integrated Product Team (IPT) concept. The IPT conceptprovides a method of managing tasks carried out across multiple organizations and locations. Each Level One WBS element is managed byan IPT responsible for delivering the required sub-systems on time, within the specified cost and meeting the project requirements.The ALMAProject Control Managenent System (PMCS) is centralized under the leadership of the ALMA Project Controller located in the ALMA offices inSantiago, Chile with staff located in Chile, Europe and North America.Purpose and scope of the position: The successful candidate is part of the ALMA PMCS team and reports to the European Project Con-troller. Responsibilities include: maintaining integrated project schedules for activities conducted in Europe; coordinating work activities withIPT’s; progressing schedules and assisting in identifying and resolving schedule conflicts; providing Earned Value Management (EVM) report-ing and variance analysis in conjunction with the Project Controller; developing and controlling associated project documentation; providingchange and problem management.The position involves duty travels to ALMA partner regions, Chile, Europe and North America.Duties and responsibilities:

– Using Project Planning & Control (PP&C) tools, document and maintain master and sub-project plans including schedule, tasks, mile-stones, resource assignments and time/expense tracking for large, geographically dispersed, IPT activities and subcontracted work.

– Configure and customize PP&C software programs including determining user requirements; establishing and maintaining multiple remoteuser accounts; establishing data interfaces; and developing internal and executive reports.

– Assist the Project Controller with analyzing and reporting on the required master and sub-project plans to plan, track and display IPTefforts and performance.

– Assist with the integration of cost accounting and other plans into the master project schedule to facilitate EVM; collect periodic updatesand reconcile differences; develop, analyze and report on EVM metrics as directed by the Management IPT.

– Assist with development and maintenance of additional project management tools and processes to assist the Project Controller in over-seeing the IPT efforts and activities.

Professional requirements/qualifications:– University degree in Engineering or technology related fields.– Significant experience using medium to high-end Project Planning & Control tools (e.g. Open Plan, COBRA, Primavera, MS Project Serv-

er, MS Project Professional, etc.).– Experience of working in large collaborative multi-cultural project environments.– Ability to coach technical and financial personnel in Earned Value Management (EVM) and project scheduling methods (e.g., critical path

method).– Project Management Professional (PMP) certification is a plus.– Proficiency in MS Office applications (Outlook, Word, PowerPoint, Excel and Access).– Multi-tasking competencies to manage multiple efforts or projects.– Excellent communication skills, a good command of the English language and a strong sense of team spirit are essential.

Duty station: Garching near Munich, Germany.Starting date: As soon as possible.

see next page

Managerial Competence Criteria – Ability to identify key strategic issues, opportunities and risks;– Capability to communicate links between the Organization's strategy and the work unit's goals; establish/identify and communicate broad

and compelling organizational directions;– Strong managerial/leadership skills; with demonstrated experience to perform and/or oversee the analysis of complex human resources,

financial, logistical or administrative management policy and program issues;– Ability to effectively lead, supervise, mentor, develop and evaluate staff and design training/skills enhancement initiatives to ensure effec-

tive staff development;– Excellent oral/written communication skills in English. Working knowledge of Spanish or the willingness to learn it;– Proficiency with the Microsoft Office package.

Remuneration and contract: We offer an attractive remuneration package including a competitive salary (tax-free), comprehensive pensionscheme and medical, educational and other social benefits as well as financial support in relocating your family. The initial contract is for aperiod of three years with the possibility of a fixed-term extension or permanence. Staff Category: International Staff Member. The grade may be subject to change according to qualifications and the number of years of expe-rience.Duty Station/Working Schedule/Place of residence: The ESO Paranal Observatory, at Cerro Paranal (120 km south of Antofagasta), is locat-ed at an altitude of 2600 m. The working system is typically 5/2 (Monday to Friday) or 8/6 with accommodation provided on site. After a firstperiod of 5/2 the final turno will be agreed upon with the supervisor according to the operational requirements. The place of residence will beAntofagasta or Santiago.Starting date: As soon as possible.Applications: If you are interested in working in a stimulating international research environment and in areas of frontline science and tech-nology, please send us your CV in English and the ESO Application Form (to be obtained from the ESO Home Page at http://www.eso.org/gen-fac/adm/pers/forms/) by

30 April 2005.

For further information please contact Mr. Walter Demartis at [email protected] are also strongly encouraged to consult the ESO Home Page (http://www.eso.org) for additional information.

Although preference will be given to nationals of the Member States of ESO: Belgium, Denmark, Finland, France, Germany, Italy, The Nether-lands, Portugal, Sweden, Switzerland, and United Kingdom, no nationality is à priori excluded. The post is equally open to suitably qualifiedmale and female applicants.


PPERSONNELERSONNEL MMOVEMENTSOVEMENTS

1 December 2004 – 28 February 2005

ArrivalsEurope

Araujo Hauck, Constanza (RCH) Paid AssociateBoxheimer, Jutta (D) Graphics DesignerDelmotte, Nausicaa (F) Paid AssociateDemartis, Walter (I) Personnel OfficerFechner, Matthias (D) StudentHarrison, Paul (UK) Paid AssociateHastie, Morag Ann (UK) StudentKiupel, Katarina (D) ERP System SpecialistNilsson, Kim (S) StudentRahoui, Farid (F) StudentRobinson, Mark (UK) Programme ControllerStrazzullo, Veronica (I) StudentSurdej, Isabelle (B) Paid Associate

Chile

Amado Gonzalez, Pedro Jose (E) Operations AstronomerArenas, Eduardo (PE) Procurement OfficerCamuri, Massimiliano (I) Electronics EngineerCarstens, Johan (S) Electronics Engineerde Brito Leal, Luis Filipe (P) StudentGonzalez, Victor (RCH) Software EngineerHorst, Hannes (D) StudentMarkar, Kiriako (RCH) Logistics SupervisorNicoud, Jean-Luc (CH) Mechanical EngineerReveco, Johnny (RCH) Software Engineer

DeparturesEurope

Beckers, Jean-Louis (B) Project ControllerCioni, Maria-Rosa (I) FellowMackowiak, Bernhard (D) Paid AssociateHuxley, Alexis (UK) Software EngineerPignata, Giuliano (I) Student

Chile

Alloin, Danielle (F) Head of Science VitacuraArredondo, Diego (RCH) System AdministratorBilleres, Malvina (F) FellowJohansson, Lars Erik (S) Paid AssociatePantin, Eric (F) Paid AssociateRagaini, Silvia (I) StudentNakos, Theodoros (GR) StudentCasquilho Faria, Daniel (S) StudentSbordone, Luca (I) StudentSuc, Vincent (F) Student

Remuneration and contract: We offer an attractive remuneration package including a competitive salary (tax free), comprehensive pensionscheme and medical, educational and other social benefits as well as financial support in relocating your family. The initial contract is for aperiod of three years with the possibility of a fixed-term extension or permanence. The title, grade and level of responsibility may be subjectto change according to qualification and the number of years of experience. Serious consideration will be given to outstanding candidateswilling to be seconded to ESO on extended leaves from their home institutions.Applications: If you are interested in working in a stimulating international research environment and in areas of frontline science and tech-nology, please send us your CV (in English) and the ESO Application Form (http://www.eso.org/gen-fac/adm/pers/forms/) together with thenames of four individuals willing to provide professional reference letters. Applications should be submitted by

15 April 2005.

For further information please contact Mr. Roland Block, Head of Personnel Department, Tel.: +49-89-3200-6589. You are also strongly encour-aged to consult the ESO Home Page (http://www.eso.org/) for additional information about ESO.


Recent PrRecent Proceedings froceedings from the ESOom the ESO AstrAstrophysics Symposiaophysics Symposia

Volume

16/2003

15/2004

14/2003

13/2003

12/2003

11/2003

10/2003

9/2003

8/2003

7/2003

6/2003

5/2003

Title

Astronomy, Cosmology and Fundamental Physics

Toward an International Virtual Observatory

Extragalactic Globular Cluster Systems

From Twilight to Highlight: The Physics of Supernovae

The Mass of Galaxies at Low and High Redshift

Lighthouses of the Universe: The Most Luminous Celes-tial Objects and Their Use for Cosmology

Scientific Drivers for ESO Future VLT/VLTI Instrumentation

The Origin of Stars and Planets: The VLT View

Gamma-Ray Bursts in the Afterglow Era

Deep Fields

Mining the Sky

Quasars, AGNs and Related Research Across 2000

For further information look at http://www.springerlink.com/openurl.asp?genre=journal&issn=1431-2433

Editors

Peter A. Shaver, Luigi DiLella, Alvaro Giménez

Peter J. Quinn, Krzysztof M. Górski

Markus Kissler-Patig

Wolfgang Hillebrandt, Bruno Leibundgut

Ralf Bender, Alvio Renzini

Marat Gilfanov, Rashid Sunyeav,Eugene Churazov

Jaqueline Bergeron, Guy Monnet

João F. Alves, Mark J. McCaughrean

Enrico Costa, Filippo Frontera, Jens Hjorth

Stefano Cristiani, Alvio Renzini, Robert E. Williams

Anthony J. Banday, Saleem Zaroubi,Matthias Bartelmann

Giancarlo Setti, Jean-Pierre Swings


MESSENGER INDEX 2004 (NMESSENGER INDEX 2004 (NOSOS. 11. 1155 –118)–118)

S U B J E C T I N D E XS U B J E C T I N D E X

C. Cesarsky: ESO Welcomes Finland asEleventh Member State 117, 2

K. Mattila, M. Tornikoski, I. Tuominen, E. Valtaoja: Astronomy in Finland 117, 3

TELESCOPES AND INSTRUMENTAION

C. Cesarsky, R. Giacconi, R. Gilmozzi: FiveYears VLT 115, 2

A. Renzini: Excerpts from the First Five Years of VLT Science (1999–2004) 115,5

A. Moorwood, S. D’Odorico: New VLTInstruments Underway 115, 8

R. Arsenault, N. Hubin, M. Le Louarn, G. Monnet, M. Sarazin: Towards an Adap-tive Secondary for the VLT? 115, 11

B. Koehler, the AT1 Assembly and Commis-sioning Team: The First Light of the FirstVLTI Auxiliary Telescope 115, 15

F. Patat: Night sky brightness during sunspotmaximum at Paranal 115, 18

A. Richichi, R. G. Petrov: Under the sign ofthree: AMBER joins the VLTI 116, 2

P. Ballester, I. Percheron, C. Sabet, T. Licha,D. J. McKay, S. Morel, M. Petr-Gotzens,A. Richichi, M. VanDenAncker, M.Wittkowski: The MIDI Data Flow: FirstObserving Period 116, 4

G. Mathys: Five years of science operationsof the VLT on Paranal 116, 8

R. Gilmozzi, J. Melnick: Th’ESObservatoire:Merging La Silla and Paranal Observato-ries 116, 13

E. Pompei, M. Billeres, T. Dall, C. Foellmi,O. Hainaut, L. Germany, G. Lo Curto, I. Saviane: Do it your-self: La Silla quick-look tools 116, 16

L. Germany: News from the La Silla Site 116,17

P. O. Lagage, J. W. Pel, M. Authier, J. Belorgey, A. Claret, C. Doucet, D. Dubreuil, G. Durand, E. Elswijk, P. Girardot, H. U. Käufl, G. Kroes, M. Lortholary, Y. Lussignol, M. Marchesi,E. Pantin, R. Peletier, J.-F. Pirard, J. Pragt,Y. Rio, T. Schoenmaker, R. Siebenmorgen,A. Silber, A. Smette, M. Sterzik, C. Veyssiere: Successful Commissioningof VISIR: The mid-infrared VLT Instru-ment 117, 12

H. Bonnet, R. Abuter, A. Baker,W. Bornemann, A. Brown, R. Castillo, R. Conzelmann, R. Damster, R. Davies, B. Delabre, R. Donaldson, C. Dumas, F. Eisenhauer, E. Elswijk, E. Fedrigo, G. Finger, H. Gemperlein, R. Genzel, A. Gilbert, G. Gillet, A. Goldbrunner, M. Horrobin, R. ter Horst, S. Huber,

N. Hubin, C. Iserlohe, A. Kaufer, M. Kissler-Patig, J. Kragt, G. Kroes, M. Lehnert, W. Lieb, J. Liske, J.-L. Lizon,D. Lutz, A. Modigliani, G. Monnet, N. Nesvadba, J. Patig, J. Pragt, J. Reunanen, C. Röhrle, S. Rossi, R. Schmutzer, T. Schoenmaker, J. Schreiber, S. Ströbele, T. Szeifert, L. Tacconi, M. Tecza, N. Thatte, S. Tordo,P. van der Werf, H.Weisz: First Light ofSINFONI at the VLT 117, 17

R. Arsenault, R. Donaldson, C. Dupuy, E. Fedrigo, N. Hubin, L. Ivanescu, M. Kasper,S. Oberti, J. Paufique, S. Rossi,A. Silber, B. Delabre, J.-L. Lizon, P. Gigan: Third MACAO-VLTI CurvatureAdaptive Optics System now installed117, 25

J. P. Emerson, W. J. Sutherland, A. M. McPherson, S. C. Craig, G. B.Dalton, A. K. Ward: The Visible & InfraredSurvey Telescope for Astronomy 117, 27

T. Wilson: ALMA News 117, 32L. Germany: La Silla News 117, 32C. Izzo, N. Kornweibel, D. McKay, R. Palsa,

M. Peron, M. Taylor: Gasgano & ESOVIMOS Pipeline Released 117, 33

D. R. Silva, M. Peron: VLT Science ProductsProduced by Pipelines: A Status Report118, 2

A. Modigliani, G. Mulas, I. Porceddu, B. Wolff, F. Damiani, K. Banse: TheFLAMES-UVES Pipeline 118, 8

F. Patat: Observing During Bright Time: Tipsand Tricks 118, 11

J. Alves, M. Lombardi: The Sky Distributionof VLT Observations 118, 15

T. Wilson, C. De Breuck, M. Zwaan: AProgress Report on ALMA 118, 16

G. H. Tan, B. D. Jackson, B. Lazareff, J. Adema, A. M. Baryshev, R. Hesper, T. M. Klapwijk, M. Kroug, S. Mahieu, D. Maier, K. Schuster, K. Wielinga, T. Zijlstra: The European Receivers forALMA 118, 18

REPORTS FROM OBSERVERS

M. R. L. Cioni, H. J. Habing, C. Loup, N. Epchtein, E. Deul: DENIS results onthe Magellanic Clouds 115, 22

M. D. Lehnert, M. Bremer: High RedshiftGalaxies and the Sources of Reionization115, 27

E. Tolstoy, M. Irwin, A. Cole, F. Fraternali,T. Szeifert, G. Marconi: Stellar Spectro-scopy of Individual Stars in Local GroupGalaxies with the VLT: Kinematics andCalcium Triplet abundances 115, 32

T. Encrenaz, E. Lellouch, P. Drossart, H. Feuchtgruber, G. S. Orton, S. K. Atreya:First Detection of Carbon Monoxide in theAtmosphere of Uranus 115, 35

A. Koch, E. K. Grebel, M. Odenkirchen, J. A. R. Caldwell: Correcting spatialgra-dients: The Case of Wide Field ImagerPhotometry 115, 37

I.Appenzeller, R. Bender, A. Böhm, S. Frank, K. Fricke, A. Gabasch, J. Heidt, U. Hopp,K. Jäger, D. Mehlert, S. Noll, R. Saglia, S. Seitz, C. Tapken, B. Ziegler: ExploringCosmic Evolution with the FORS DeepField 116, 18

ESO PR Photo 20/04: Revisiting the OrionNebula, Wide Field Imager Provides NewView of a Stellar Nursery 116, 24

R. Srianand, P. Petitjean, H. Chand, B. Aracil:Constraining the time variation of the finestructure constant 116, 25

E. Galliano, D. Alloin: Extragalactic Embed-ded Clusters: Exploring Star Formation116, 29

C. Melo, F. Bouchy, F. Pont, N. C. Santos, M. Mayor, D. Queloz, S. Udry: Two NewVery Hot Jupiters in the FLAMES spot-light 116, 32

ESO Press Photos 08a-c/04 and ESO PressRelease 09/04: New VLT/NACO Imagesof Titan 116, 35

R. Chini, V. Hoffmeister, S. Kimeswenger,M. Nielbock, D. Nürnberger, L. Schmitdtobreick, M. Sterzik: The Birthof a Massive Star 117, 36

N. Christlieb, D. Reimers, L. Wisotzki: TheStellar Content of the Hamburg/ESO Sur-vey 117, 40

C. Cacciari, A. Bragaglia, E. Carretta:FLAMES Spectroscopy of RGB stars inthe Globular Cluster NGC 2808: MassLoss and Na-O Abundances 117, 47

L. Pasquini, P. Bonifacio, S. Randich, D. Galli, R. G. Gratton: VLT Observationsof Beryllium in a Globular Cluster: Aclockfor the early Galaxy and new insights intoGlobular Cluster formation 117, 50

P. Kervella, D. Bersier, N. Nardetto, D. Mourard, P. Fouqué, V. Coudé duForesto: Cepheid Pulsations Resolved bythe VLTI 117, 53

P. Padovani, M. G. Allen, P. Rosati, N. A. Walton: First Science for the Virtu-al Observatory 117, 58

C. Lidman: Observing Distant Type Ia Super-novae with the ESO VLT 118, 24

ESO PR Photo 28e/04: ESO Views of Earth-Approaching Asteroid Toutatis 118, 30

M. Della Valle, D. Malesani, G. Chincarini,


L. Stella, G. Tagliaferri, L. A. Antonelli,S. Campana, S. Covino, F. Fiore, N. Gehrels, K. Hurley, L. J. Pellizza, F. M. Zerbi, L. Angelini, L. Burderi, D. Burrows, M. Capalbi, P. Caraveo, E. Costa, G. Cusumano, P. Filliatre, D. Fugazza, R. Gilmozzi, P. Giommi, P. Goldoni, G. L. Israel, E. Mason, K. Mason, A. Melandri, S. Mereghetti, I. F. Mirabel, E. Molinari, A. Moretti, J. Nousek, P. O’Brien, J. Osborne, R. Perna, M. Perri, L. Piro, E. Puchnarewicz, M. Vietri (the MISTICICollaboration): Supernovae Shed Lighton Gamma-Ray Bursts 118, 31

P. Vreeswijk, P. Møller, C. Ledoux, S. Ellison, N. Masetti, J. Fynbo, P. Jakobsson, J. Hjorth: GRB Afterglows:Illuminating the Star-Forming Universe118, 35

J. Bergeron, P. Petitjean, B. Aracil, C. Pichon,E. Scannapieco, R. Srianand, P. Boissé, R. F. Carswell, H. Chand, S. Cristiani, A. Ferrara, M. Haehnelt, A. Hughes, T.-S. Kim, C. Ledoux, P. Richter, M. Viel:The Large Programme “Cosmic Evolutionof the IGM” 118, 40

E. Vanzella, S. Cristiani, M. Dickinson, H. Kuntschner, L. A. Moustakas, M. Nonino, P. Rosati, D. Stern, C. Cesarsky, S. Ettori, H. C. Ferguson, R. A. E. Fosbury, M. Giavalisco, J. Haase,A. Renzini, A. Rettura, P. Serra and the GOODS Team: VLT/FORS2 Spec-troscopy in the GOODS-South Field 118,45

A. Cimatti, E. Daddi, A. Renzini, P. Cassata,E. Vanzella, L. Pozzetti, S. Cristiani, A. Fontana, G. Rodighiero, M. Mignoli,G. Zamorani: Unveiling Old MassiveSpheroidal Galaxies in the Young Uni-verse 118, 51

R. Cayrel, M. Spite: The First Stars: What Weknow and Do Not Know 118, 55

B. Nordström, M. Mayor, J. Andersen, J. Holmberg, B. R. Jørgensen, F. Pont, E. H. Olsen, S. Udry, N. Mowlavi: A Live-lier Picture of the Solar Neighbourhood118, 61

OTHER ASTRONOMICAL NEWS

H. Van Winckel: The Users’ Committee 115,40

S. Wagner, B. Leibundgut: Report on the ESOWorkshop on Large Programmes and Sur-veys 115, 41

I. Corbett: A Report on a Workshop on FutureLarge-Scale Projects and Programmes in

Astronomy and Astrophysics 115, 43ESO Press Release 02/04: Finland to Join

ESO 115, 44D. Alloin, D. Minniti: Report on the first

Advanced Chilean School on Astro-physics “Extrasolar Planets and BrownDwarfs” 115, 45

D. Alloin: Report on the International Work-shop “Physics of Active Galactic Nucleiat all Scales” 115, 46

C. Hummel, D. Alloin: Report on a Micro-Workshop held at ESO/Vitacura on 2004 January 28 “Optical InterferometryBrings New Frontiers in Astrophysics”115, 47

Fellows at ESO: Nicolas Cretton,Emanuel Daddi, Poshak Gandhi, Dieter Nürnberger 115, 47

R. West, H. Boffin: The Venus Transit 2004(VT-2004) Programme 115, 49

H. Boffin, R. West: Venus Transit 2004: A Tremendous Success! 116, 39

C. Madsen: Serving European Science: TheEIROforum Collaboration 116, 42

D. Alloin, E. Mason, K. O’Brian: Report onThe visit of Prof. R. Sunyaev to ESO/Chileand the Topical Meeting “Accretion ontoCompact Objects” 116, 44

J. R. Walsh, M. Rejkuba: Report on the ESOConference: Planetary Nebulae beyondthe Milky Way 116, 44

Fellows at ESO: Thomas Dall, Marina Rejkuba, Gijsbert Verdoes, Paul Vreeswijk 116, 46

Former ESO DG Adriaan Blaauw turned 90116, 47

E. Janssen: ESO at the Exploring the Fron-tier Symposium 116, 47

ESO Press Release 23/04: Is this Speck ofLight an Exoplanet? 117, 11

D. Alloin: Students at ESO/Santiago: ThePeriod 1999-2004 117, 61

A. Merloni, S. Nayakshin: Report on the jointMPA/MPE/ESO/USM workshop onGrowing Black Holes: Accretion in a Cos-mological Context 117, 62

H. Boffin, R. West: Report on a brainstorm-ing meeting about the ALMA Interdisci-plinary Teaching Project 117, 63

D. H. Lorenzen: Obituary Jürgen Stock1923–2004 117, 65

C. Madsen: Report on the EuroScience OpenForum 2004 117, 66

Fellows at ESO: Carlos de Breuck, PierreKervella, Celine Peroux 117, 67

S. Randich, L. Pasquini: Report on the ESO-Arcetri Workshop on Chemical Abun-dances and Mixing in Stars in the Milky

Way and its Satellites 118, 66T. L. Wilson, E. F. van Dishoeck: ALMA

Community Day 118, 67 M. Dennefeld: The NEON School Enters a

New Era 118, 68D. Alloin, C. Lidman: Report on the 2004

IAOC Conference “The Cool Universe:Exploring Cosmic Dawn” 118, 69

H. Boffin, R. West: The “Venus Transit Expe-rience” 118, 69

Personnel Movements 118, 70C. Madsen, S. D’Odorico: ESO Presentation

in Copenhagen 118, 71

ANNOUNCEMENTS

B. Pirenne: Upcoming Change in the Sup-ported Archive Output Media 115, 52

ESO-Acetri Conference on Chemical Abun-dances and Mixing in Stars in the MilkyWay and its Satellites 115, 52

Personnel Movements 115, 53ESO Studentship Programme 115, 53ESO Vacancy: ESO Representative in Chile,

Head of the Office for Science Santiago115, 54

ESO Vacancy: Operations Staff Astronomer,Deputy Head of the Paranal Science Oper-ations Department 115, 54

ESO Vacancy: Staff Astronomer 115, 55ESO Vacancy: Operations Staff Astronomer

(VLTI) 115, 55B. Pirenne, P. Quinn: Science Products from

large programmes to be available from theESO archive 116, 48

C. Madsen: Visit EIROforum at ESOF 2004116, 48

NOVA-ESO-MPIAworkshop on VLTI/MIDIData reduction, Analysis and Science 116,49

International Astronomical Observatories inChile (IAOC) Workshop in 2004 “TheCool Universe: Observing Cosmic Dawn” 116, 49

ESO Vacancy: Senior Astronomer, Directorof the La Silla/Paranal Observatories 116,50

Personnel Movements 116, 50ESO Fellowship Progamme 2004/2005 116,

51A. Renzini: A New Start for Public Surveys

117, 68Personnel Movements 117, 69ESO Vacancy: Desktop Publishing/Graphics

Specialist 117, 70ESO Workshop Proceedings Still Available

117, 70ESO Fellowship Progamme 2004/2005 117,

71


AA

D. Alloin, D. Minniti: Report on the firstAdvanced Chilean School on Astro-physics “Extrasolar Planets and BrownDwarfs” 115, 45

D. Alloin: Report on the International Work-shop “Physics of Active Galactic Nucleiat all Scales” 115, 46

D. Alloin, E. Mason, K. O’Brian: Report onThe visit of Prof. R. Sunyaev to ESO/Chileand the Topical Meeting “Accretion ontoCompact Objects” 116, 44

D. Alloin: Students at ESO/Santiago: ThePeriod 1999-2004 117, 61

D. Alloin, C. Lidman: Report on the 2004IAOC Conference “The Cool Universe:Exploring Cosmic Dawn” 118, 69

J. Alves, M. Lombardi: The Sky Distributionof VLT Observations 118, 15

I. Appenzeller, R. Bender, A. Böhm, S. Frank, K. Fricke, A. Gabasch, J. Heidt, U. Hopp,K. Jäger, D. Mehlert, S. Noll, R. Saglia, S. Seitz, C. Tapken, B. Ziegler: ExploringCosmic Evolution with the FORS DeepField 116, 18

R. Arsenault, N. Hubin, M. Le Louarn, G. Monnet, M. Sarazin: Towards an Adap-tive Secondary for the VLT? 115, 11

R. Arsenault, R. Donaldson, C. Dupuy, E. Fedrigo, N. Hubin, L. Ivanescu, M. Kasper, S. Oberti, J. Paufique, S. Rossi,A. Silber, B. Delabre, J.-L. Lizon, P. Gigan: Third MACAO-VLTI CurvatureAdaptive Optics System now installed117, 25

BB

P. Ballester, I. Percheron, C. Sabet, T. Licha,D. J. McKay, S. Morel, M. Petr-Gotzens,A. Richichi, M. VanDenAncker, M.Wittkowski: The MIDI Data Flow: FirstObserving Period 116, 4

J. Bergeron, P. Petitjean, B. Aracil, C. Pichon,E. Scannapieco, R. Srianand, P. Boissé, R. F. Carswell, H. Chand, S. Cristiani, A. Ferrara, M. Haehnelt, A. Hughes, T.-S. Kim, C. Ledoux, P. Richter, M. Viel:The Large Programme “Cosmic Evolutionof the IGM” 118, 40

H. Boffin, R. West: Venus Transit 2004: A Tremendous Success! 116, 39

H. Boffin, R. West: Report on a brainstorm-ing meeting about the ALMA Interdisci-plinary Teaching Project 117, 63

H. Boffin, R. West: The “Venus Transit Expe-rience” 118, 69

H. Bonnet, R. Abuter, A. Baker, W. Bornemann, A. Brown, R. Castillo, R. Conzelmann, R. Damster, R. Davies,

A U T H O R I N D E XA U T H O R I N D E X

B. Delabre, R. Donaldson, C. Dumas, F. Eisenhauer, E. Elswijk, E. Fedrigo, G. Finger, H. Gemperlein, R. Genzel, A. Gilbert, G. Gillet, A. Goldbrunner, M. Horrobin, R. ter Horst, S. Huber, N. Hubin, C. Iserlohe, A. Kaufer, M. Kissler-Patig, J. Kragt, G. Kroes, M. Lehnert, W. Lieb, J. Liske, J.-L. Lizon,D. Lutz, A. Modigliani, G. Monnet, N. Nesvadba, J. Patig, J. Pragt, J. Reunanen, C. Röhrle, S. Rossi, R. Schmutzer, T. Schoenmaker, J. Schreiber, S. Ströbele, T. Szeifert, L. Tacconi, M. Tecza, N. Thatte, S. Tordo,P. van der Werf, H.Weisz: First Light ofSINFONI at the VLT 117, 17

CC

C. Cacciari, A. Bragaglia, E. Carretta:FLAMES Spectroscopy of RGB stars inthe Globular Cluster NGC 2808: MassLoss and Na-O Abundances 117, 47

R. Cayrel, M. Spite: The First Stars: What Weknow and Do Not Know 118, 55

C. Cesarsky, R. Giacconi, R. Gilmozzi: FiveYears VLT 115, 2

C. Cesarsky: ESO Welcomes Finland asEleventh Member State 117, 2

R. Chini, V. Hoffmeister, S. Kimeswenger,M. Nielbock, D. Nürnberger, L. Schmitdtobreick, M. Sterzik: The Birthof a Massive Star 117, 36

N. Christlieb, D. Reimers, L. Wisotzki: TheStellar Content of the Hamburg/ESO Sur-vey 117, 40

A. Cimatti, E. Daddi, A. Renzini, P. Cassata,E. Vanzella, L. Pozzetti, S. Cristiani, A. Fontana, G. Rodighiero, M. Mignoli,G. Zamorani: Unveiling Old MassiveSpheroidal Galaxies in the Young Uni-verse 118, 51

M. R. L. Cioni, H. J. Habing, C. Loup, N. Epchtein, E. Deul: DENIS results onthe Magellanic Clouds 115, 22

I. Corbett: A Report on a Workshop on FutureLarge-Scale Projects and Programmes inAstronomy and Astrophysics 115, 43

DD

M. Della Valle, D. Malesani, G. Chincarini, L. Stella, G. Tagliaferri, L. A. Antonelli,S. Campana, S. Covino, F. Fiore, N. Gehrels, K. Hurley, L. J. Pellizza, F. M. Zerbi, L. Angelini, L. Burderi, D. Burrows, M. Capalbi, P. Caraveo, E. Costa, G. Cusumano, P. Filliatre, D. Fugazza, R. Gilmozzi, P. Giommi, P. Goldoni, G. L. Israel, E. Mason, K. Mason, A. Melandri, S. Mereghetti,

I. F. Mirabel, E. Molinari, A. Moretti, J. Nousek, P. O’Brien, J. Osborne, R. Perna, M. Perri, L. Piro, E. Puchnarewicz, M. Vietri (the MISTICICollaboration): Supernovae Shed Lighton Gamma-Ray Bursts 118, 31

M. Dennefeld: The NEON School Enters aNew Era 118, 68

EE

J. P. Emerson, W. J. Sutherland, A. M.McPherson, S. C. Craig, G. B. Dalton, A. K. Ward: The Visible & Infrared Sur-vey Telescope for Astronomy 117, 27

T. Encrenaz, E. Lellouch, P. Drossart, H. Feuchtgruber, G. S. Orton, S. K. Atreya:First Detection of Carbon Monoxide in theAtmosphere of Uranus 115, 35

GG

E. Galliano, D. Alloin: Extragalactic Embed-ded Clusters: Exploring Star Formation116, 29

L. Germany: News from the La Silla Site 116,17

L. Germany: La Silla News 117, 32R. Gilmozzi, J. Melnick: Th’ESObservatoire:

Merging La Silla and Paranal Observato-ries 116, 13

HH

C. Hummel, D. Alloin: Report on a Micro-Workshop held at ESO/Vitacura on 2004 January 28 “Optical InterferometryBrings New Frontiers in Astrophysics”115, 47

II

C. Izzo, N. Kornweibel, D. McKay, R. PalsaM. Peron, M. Taylor: Gasgano & ESOVIMOS Pipeline Released 117, 33

JJ

E. Janssen: ESO at the Exploring the Fron-tier Symposium 116, 47

KK

P. Kervella, D. Bersier, N. Nardetto, D. Mourard, P. Fouqué, V. Coudé duForesto: Cepheid Pulsations Resolved bythe VLTI 117, 53


A. Koch, E. K. Grebel, M. Odenkirchen, J. A. R. Caldwell: Correcting spatialgra-dients: The Case of Wide Field ImagerPhotometry 115, 37

B. Koehler, the AT1 Assembly and Commis-sioning Team: The First Light of the FirstVLTI Auxiliary Telescope 115, 15

LL

P. O. Lagage, J. W. Pel, M. Authier, J. Belorgey, A. Claret, C. Doucet, D. Dubreuil, G. Durand, E. Elswijk, P. Girardot, H. U. Käufl, G. Kroes, M. Lortholary, Y. Lussignol, M. Marchesi,E. Pantin, R. Peletier, J.-F. Pirard, J. Pragt, Y. Rio, T. Schoenmaker, R. Siebenmorgen, A. Silber, A. Smette, M. Sterzik, C. Veyssiere: Successful Com-missioning of VISIR: The mid-infraredVLT Instrument 117, 12

M. D. Lehnert, M. Bremer: High RedshiftGalaxies and the Sources of Reionization115, 27

C. Lidman: Observing Distant Type Ia Super-novae with the ESO VLT 118, 24

D. H. Lorenzen: Obituary Jürgen Stock1923–2004 117, 65

MM

C. Madsen: Serving European Science: TheEIROforum Collaboration 116, 42

C. Madsen: Visit EIROforum at ESOF 2004116, 48

C. Madsen: Report on the EuroScience OpenForum 2004 117, 66

C. Madsen, S. D’Odorico: ESO Presentationin Copenhagen 118, 71

G. Mathys: Five years of science operationsof the VLT on Paranal 116, 8

K. Mattila, M. Tornikoski, I. Tuominen, E. Valtaoja: Astronomy in Finland 117, 3

C. Melo, F. Bouchy, F. Pont, N. C. Santos, M. Mayor, D. Queloz, S. Udry: Two NewVery Hot Jupiters in the FLAMES spot-light 116, 32

A. Merloni, S. Nayakshin: Report on the joint MPA/MPE/ESO/USM workshop onGrowing Black Holes: Accretion in a Cos-mological Context 117, 62

A. Modigliani, G. Mulas, I. Porceddu, B. Wolff, F. Damiani, K. Banse: TheFLAMES-UVES Pipeline 118, 8

A. Moorwood, S. D’Odorico: New VLTInstruments Underway 115, 8

NN

B. Nordström, M. Mayor, J. Andersen, J. Holmberg, B. R. Jørgensen, F. Pont, E. H. Olsen, S. Udry, N. Mowlavi: A Live-lier Picture of the Solar Neighbourhood118, 61

PP

P. Padovani, M. G. Allen, P. Rosati, N. A. Walton: First Science for the Virtu-al Observatory 117, 58

L. Pasquini, P. Bonifacio, S. Randich, D. Galli, R. G. Gratton: VLT Observationsof Beryllium in a Globular Cluster: Aclockfor the early Galaxy and new insights intoGlobular Cluster formation 117, 50

F. Patat: Night sky brightness during sunspotmaximum at Paranal 115, 18

F. Patat: Observing During Bright Time: Tipsand Tricks 118, 11

B. Pirenne: Upcoming Change in the Sup-ported Archive Output Media 115, 52

B. Pirenne, P. Quinn: Science Products fromlarge programmes to be available from theESO archive 116, 48

E. Pompei, M. Billeres, T. Dall, C. Foellmi,O. Hainaut, L. Germany, G. Lo Curto, I. Saviane: Do it your-self: La Silla quick-look tools 116, 16

RR

S. Randich, L. Pasquini: Report on the ESO-Arcetri Workshop on Chemical Abun-dances and Mixing in Stars in the MilkyWay and its Satellites 118, 66

A. Renzini: Excerpts from the First FiveYears of VLT Science (1999–2004) 115,5

A. Renzini: A New Start for Public Surveys117, 68

A. Richichi, R. G. Petrov: Under the sign ofthree: AMBER joins the VLTI 116, 2

SS

D. R. Silva, M. Peron: VLT Science ProductsProduced by Pipelines: A Status Report118, 2

R. Srianand, P. Petitjean, H. Chand, B. Aracil:Constraining the time variation of the finestructure constant 116, 25

TT

G. H. Tan, B. D. Jackson, B. Lazareff, J. Adema, A. M. Baryshev, R. Hesper, T. M. Klapwijk, M. Kroug, S. Mahieu, D. Maier, K. Schuster, K. Wielinga, T. Zijlstra: The European Receivers forALMA 118, 18

E. Tolstoy, M. Irwin, A. Cole, F. Fraternali,T. Szeifert, G. Marconi: Stellar Spec-troscopy of Individual Stars in LocalGroup Galaxies with the VLT: Kinemat-ics and Calcium Triplet abundances 115,32

VV

H. Van Winckel: The Users’ Committee 115,40

E. Vanzella, S. Cristiani, M. Dickinson, H. Kuntschner, L. A. Moustakas, M. Nonino, P. Rosati, D. Stern, C. Cesarsky, S. Ettori, H. C. Ferguson, R. A. E. Fosbury, M. Giavalisco, J. Haase,A. Renzini, A. Rettura, P. Serra and theGOODS Team: VLT/FORS2 Spec-troscopy in the GOODS-South Field 118,45

P. Vreeswijk, P. Møller, C. Ledoux, S. Ellison, N. Masetti, J. Fynbo, P. Jakobsson, J. Hjorth: GRB Afterglows:Illuminating the Star-Forming Universe118, 35

WW

S. Wagner, B. Leibundgut: Report on the ESOWorkshop on Large Programmes and Sur-veys 115, 41

J. R. Walsh, M. Rejkuba: Report on the ESOConference: Planetary Nebulae beyondthe Milky Way 116, 44

R. West, H. Boffin: The Venus Transit 2004(VT-2004) Programme 115, 49

T. Wilson: ALMA News 117, 32T. Wilson, C. De Breuck, M. Zwaan:

A Progress Report on ALMA 118, 16T. Wilson, E. F. van Dishoeck: ALMA Com-

munity Day 118, 67

ESO, the European Southern Observa-tory, was created in 1962 to “... establishand operate an astronomical observato-ry in the southern hemisphere, equippedwith powerful instruments, with the aim offurthering and organising collaboration inastronomy...” It is supported by elevencountries: Belgium, Denmark, Finland,France, Germany, Italy, the Netherlands,Portugal, Sweden, Switzerland and theUnited Kingdom. ESO operates at threesites in the Atacama desert region ofChile. The Very Large Telescope (VLT), is located on Paranal, a 2 600 m highmountain approximately 130 km south ofAntofagasta. The VLT consists of four 8.2 m diameter telescopes. These tele-scopes can be used separately, or in com-bination as a giant interferometer (VLTI).At La Silla, 600 km north of Santiago deChile at 2 400 m altitude, ESO operatesseveral optical telescopes with diametersup to 3.6 m. The third site is the 5 000 mhigh Llano de Chajnantor, near San Pedrode Atacama. Here a new submillimetretelescope (APEX) is being completed, anda large submillimetre-wave array of 64 antennas (ALMA) is under develop-ment. Over 1600 proposals are madeeach year for the use of the ESO tele-scopes. The ESO headquarters are locat-ed in Garching, near Munich, Germany.This is the scientific, technical and admin-istrative centre of ESO where technicaldevelopment programmes are carried outto provide the Paranal and La Silla obser-vatories with the most advanced instru-ments. ESO employs about 320 inter-national staff members, Fellows andAssociates in Europe and Chile, andabout 160 local staff members in Chile.

The ESO MESSENGER is published fourtimes a year: normally in March, June,September and December. ESO also pub-lishes Conference Proceedings, Pre-prints, Technical Notes and other materi-al connected to its activities. PressReleases inform the media about par-ticular events. For further information,contact the ESO Education and PublicRelations Department at the followingaddress:

EUROPEAN SOUTHERN OBSERVATORY Karl-Schwarzschild-Str. 2D-85748 Garching bei München Germany Tel. +49 89 320 06-0 Telefax +49 89 320 23 62Email: [email protected]: http://www.eso.org

The ESO Messenger:Editor: Peter ShaverTechnical editor: Jutta Boxheimerhttp://www.eso.org/messenger/

Printed by Peschke DruckSchatzbogen 35D-81805 MünchenGermany

ISSN 0722-6691

Front Cover Picture: Galaxy NGC 1097

This is an almost-true colour composite based on three images made with the multi-mode VIMOS instru-ment on the 8.2 m Melipal Unit Telescope of the VLT. The observations were made on the night of Decem-ber 9–10, 2004, in the presence of the President of the Republic of Chile, Ricardo Lagos. The exposureswere taken and pre-processed by ESO Paranal Science Operation astronomers, with additional imageprocessing by Hans Hermann Heyer (ESO). More details are available in ESO Press Release 28/04.

CCONTENTSONTENTS

Scientific Strategy Planning at ESO......................................................................................2

TTELESCOPESELESCOPES ANDAND IINSTRUMENTNSTRUMENTAATIONTIONM. CASALI ET AL. – Wide Field Infrared Imaging on the VLT with HAWK-I.............................. 6ESO PRESS RELEASE 03/05 – ESO’s Two Observatories Merge.............................................8M. KASPER ET AL. – New observing modes of NACO............................................................. 9M. WITTKOWSKI ET AL. – Observing with the ESO VLT Interferometer................................... 14A. GILLIOTTE ET AL. – Improvements at the 3.6 m Telescope................................................ 18J. ALVES – Telescope Time Allocation Tool.......................................................................... 20T. WILSON – ALMA News............................................................................................................24

RREPOREPORTSTS FROMFROM OOBSERBSERVERSVERSE. PANTIN ET AL. – VISIR, a Taste of Scientific Potential....................................................... 25O. LE FÈVRE ET AL. – The VIMOS-VLT Deep Survey First Epoch observations.....................30M. WITTKOWSKI ET AL. – Recent Astrophysical Results from the VLTI...................................36J.-U. POTT ET AL. – VLTI Observations of IRS 3.......................................................................43

OOTHERTHER AASTRONOMICALSTRONOMICAL NNEWSEWSU. GROTHKOPF ET AL. – Comparison of Science Metrics among Observatories................... 45N. DELMOTTE ET AL. – The ESO Telescope Bibliography Web Interface................................50E. JANSSEN – ESO Exhibitions in Granada and Naples........................................................ 52R. WEST – EAAE Meeting at ESO Headquarters..................................................................53V. RODRÍGUEZ, F. MIRABEL – Public Affairs Highlights in Chile...............................................54Fellows at ESO – N. Bouché, E. Depagne, M. Hartung, J. Liske......................................... 55

AANNOUNCEMENTSNNOUNCEMENTSA. LAWRENCE, S. WARREN – The UKIRT Infrared Deep Sky Survey: data access..................56Symposium on Relativity, Matter and Cosmology...............................................................57ESO Workshop on Multiple Stars across the H-R Diagram.................................................57MPA/ESO/MPE/USM Joint Astronomy Conference on

Open Questions in Cosmology: The First Billion Years....................................................58ESO Conference on Science Perspectives for 3D Spectroscopy........................................58Felix Mirabel becomes ESO Representative in Chile.......................................................... 59ESO Studentship Programme.............................................................................................59Vacancies........................................................................................................................... 60Personnel Movements........................................................................................................ 63ANNUAL INDEX 2004............................................................................................................. 64

messenger-no119

Documents

Transcript of messenger-no119