N (with 5 as the default value) pointingson the sky. Although this will nearly cov-er all the gaps in the focal plane andmaximizes the sky coverage, the con-text map of such data is complex. Anadvantage is that it will be relativelyeasy to couple the photometry amongthe individual CCDs.

• Jitter has offsets matching thesmallest gaps in CCDs ∼ 5 pixels. Thismode optimizes the homogeneity of thecontext map and will be used during ob-servations for which the wide gaps arenot critical, but which, for instance, re-quire a well-mapped smoothly varyingPSF.

• Stare allows re-observing one fixedpointing position multiple times. It is themain workhorse for monitoring the in-strument and allows detection of opticaltransients.

• SSO is the mode for observingSolar System Objects, which requiresnon-sidereal tracking.

For all these modes dedicated ob-serving templates are being developed.

An observing strategy employs oneor a combination of the basic observingmodes. It also defines a number of ad-ditional instructions for scheduling ofthe observations. We distinguish thefollowing strategies:

• Standard which consists of a singleobservation (observation block)

• Deep which does deep integrations,possibly taken at selected atmosphericconditions over several nights

• Freq which frequently visits (moni-tors) the same field on time scalesranging from minutes to months andhas overriding priority on the telescopeschedule

• Mosaic maps areas of the sky larg-er than 1 degree, which is essentiallyan item for the scheduling, as thepipeline has to produce uniform qualitydata anyway. The combination of vari-ous field centres into one image is notconsidered a standard pipeline task.

The observing modes and strategiesare fully integrated with the data reduc-tion software being developed by theOmegaCAM consortium. We distin-guish between a calibration pipelineproducing and qualifying calibrationfiles, and an image pipeline that appliesthe calibration files to raw data andtransforms them into astrometricallyand photometrically calibrated images.ESO users will be provided with theoutput of the image pipeline, run inGarching, on the data contained in asingle OB. The nominal photometricaccuracy of this pipeline will be ± 0.05mag, exceptionally ± 0.01 mag. Thenominal accuracy for the astrometry is± 0.1 arcsec rms over the entire field ofview.

As part of the contract, theOmegaCAM consortium will deliversoftware modules that ESO will inte-grate into the image pipeline. In addi-tion, a project has been set up amongEuropean wide-field imaging groups toprovide a ‘wide-field imaging surveysystem’ that will combine pipeline pro-cessing of image data with archivingand data mining tools. Further detailscan be found on http://www.astro-wise.org, and in Valentijn & Kuijken(2002).

The development of the analysissoftware is being done by a team basedin Groningen and Leiden, led by E.Valentijn.

5. Current Status

The OmegaCAM project is now wellinto in the manufacturing phase. Mostof the CCDs have been delivered andtested; most of the mechanics exist andare ready to be integrated; instrumentcontrol and data analysis software isbeing coded. Extensive tests in Europeare foreseen for the second half of2003, and the camera should see firstlight early in 2004. Exciting times!

Acknowledgements

The consortium was formed in 1998in response to an announcement of op-portunity from ESO, and comprises in-stitutes in the Netherlands (NOVA, inparticular the Kapteyn Institute Gronin-gen and Leiden Observatory), Ger-many (in particular University Observa-tories of Munich, Göttingen and Bonn)and Italy (INAF, in particular Padua andNaples observatories). The ESO Op-tical Detector Team provides the detec-tor system at cost to the consortium.OmegaCAM is headed by PI K. Kuijken(Groningen and Leiden University) andco-PI’s R. Bender (Munich USM/MPE)and Cappellaro (INAF Naples/Padua),and project management is done by B.Muschielok and R. Häfner (USM).

OmegaCAM is funded by grants fromthe Dutch Organization for Research inAstronomy (NOVA), the German Fed-eral Ministry of Education, Science, Re-search and Technology (grants 05AV9MG1/7, AV9WM2/5, 05 AV2MGA/6and 05 AV2WM1/2), and the ItalianConsorzio Nazionale per l’Astronomiae l’Astrofisica (CNAA) and Istituto Na-zionale di Astrofisica (INAF), in additionto manpower and materials provided bythe partner institutes.

ReferencesArnaboldi, M., Capaccioli, M., Mancini, D.,

Rafanelli, P., Scaranella, R., Sedmak, G.and Vettolani, G. P., 1998. The Mes-senger 93, 30.

Baruffolo, A., Bortolussi, A., De Pizzol, L.2002, Proc. SPIE 4848, in press.

Nicklas, H., Harke, R., Wellem, W., Reif, K.2002, Proc. SPIE 4836-34, in press.

Reif, K., Klink, G., Müller, Ph. and Posch-mann, H. 2002 in Scientific Detectors forAstronomy, Beletic, Amico eds., Astro-physics and Space Sciences Library(Kluwer: Dordrecht), in press.

Valentijn, E.A. & Kuijken, K. 2002 in Towardan International Virtual Observatory,Quinn, P., ed., ESO Astrophysics Sym-posia Series (Springer-Verlag), in press.

18

The VLTI – 20 Months after First Fringes A. GLINDEMANN, ESO

1. Introduction

In 2002, the second year of fringes atParanal, the VLTI has made substantialprogress. The highlight was the comple-tion of the combination in pairs of all fourUnit Telescopes on September 15/16and 16/17 using a total of five differentbaselines. Only the combination MELI-PAL – YEPUN could not be provideddue to the current configuration of delaylines in the interferometric tunnel.

Of equal importance was the start ofa total of 150 hours shared risk scienceoperations with the VLTI in October.

Forty proposals from the communitywere received representing about 10%of all proposals submitted to ESO forthe VLT observatory. A summary of thefirst semester with VLTI science opera-tions will be given at the end of this se-mester. For Period 71, the shared riskscience operations became a part ofthe ESO Call for Proposals with 25 pro-posals submitted for the VLTI. A num-ber of observation preparation tools havebeen developed in collaboration withthe Jean-Marie Mariotti Centre for Inter-ferometry (JMMC) in Grenoble. Two ofthem are now available on the web

(http://www.eso.org/observing/etc/ pre-view.html). In the course of the year, allscience data between First Fringes inMarch 2001 and September 2002 havebeen released through the archive re-sulting in first scientific results whichare described in [1]–[4]. A summary ofthe first results is given in [5]. In thecontext of science operations, the re-sults of the on-going observations ofcalibrator stars are reported in [6], incollaboration with the NOVA ESO VLTIExpertise Centre (NEVEC) in Leiden.

Amongst the runners-up for achieve-ments are the integrated optics beam

combiner IONIC for VINCI, the accept-ance of the science instrument MIDIin Heidelberg in September (currentlybeing integrated at Paranal), and threenew contracts to extend the VLTI infra-structure: Delay Lines 4, 5 and 6 wereordered at Fokker Space in Leiden, andthe installation of the required rail sys-tems in the Delay Line Tunnel started inAugust. The contract for the fourthAuxiliary Telescope was signed withAMOS in Liège at the beginning ofSeptember (see [7]). The first PRIMAcontract for the delivery of the fringesensor unit was signed with ALENIASpazio, Torino, in July, and the secondcontract for the star separator unit willfollow soon. The PRIMA laser metrolo-gy system will be an in-house develop-ment. These three subsystems will formthe first phase of PRIMA.

IONIC was provided by the Observa-toire de Grenoble. IONIC is a beam com-biner for VINCI that can be used insteadof VINCI’s fibre beam combiner MONA.In IONIC, the beams are guided throughtwo silicon fibres into an integrated op-tics (IO) component where the beamcombination takes place. At the exit ofthe IO component, fibers are attachedguiding the light to the infrared detector.Since the fibre connectors are identicalto those used in MONA, the optical in-tegration was just a matter of decon-necting and connecting a few fibres.Unlike the K-band beam combinerMONA, IONIC, having silicon fibres andsubstrates, works in the H-band. TheVLTI had first fringes with IONIC onJuly 18, producing an interferometrictransfer function above 80% with anaccuracy of 1% which is as good asMONA (see [8]).

Last but not least there is alsoprogress with the adaptive optics sys-tems for the VLTI. The Paranalization ofthe tip-tilt sensors STRAP on two UnitTelescopes is finished, and MACAO,the high-order adaptive optics system(see [9]), saw first light in the laborato-ry in Garching in August, closing theloop on an artificial star.

2. Four Eyes Are Better

Why should one attempt to use morethan two telescopes if the interferomet-ric instrument can only combine twobeams? First of all, this exercisedemonstrated the ability to provide mul-tiple beams in the VLTI laboratory with-out losing time for mirror realignment orfor other reconfigurations. In 2003, AM-BER, the near infrared science instru-ment, will see first fringes. Then we willhave an instrument combining up tothree telescopes simultaneously.

In addition, multiple baselines are thekey to efficient interferometric imaging.The result of the observations of thestar Achernar in September, shown inFigure 1, gives an impression of the an-gular resolution capability of the VLTI

with the Unit Telescopes. The intensitydistribution was reconstructed from themeasurements with individual baselinesby Fourier transforming the distributionof values displayed in Figure 2. Everyballoon in that figure represents the re-sult of a visibility measurement with onebaseline vector. It is important to prop-erly refer to baselines as baseline vec-tors in order to understand why the dis-tribution of the balloons is considerablydifferent from the distribution of UnitTelescopes at Paranal (see Fig. 3). Thevisibility function depends on the differ-ence of telescope coordinates (project-ed perpendicular to the viewing direc-tion of the star). For instance, two tele-scopes separated by 100 m along thenorth-south direction have exactly thesame baseline vector as two other tele-scopes a few kilometres away that arealso separated by 100 m along north-south.

The baseline vectors are displayed inthe ‘uv-plane’, where the measuredvisibility forms the function value at thecoordinate given by the length andthe orientation of the baseline vector.Since the visibilities are derived fromthe intensity, a real and positive quanti-ty, the visibility function is point sym-metric; one visibility measurement de-

livers two function values, at (u, v) andat (-u, -v).

The balloons representing the meas-ured visibilities in the uv-plane aresmall disks although a point would beexpected when observing with an indi-vidual baseline. The reason is that thetelescope is not a pinhole but an 8-maperture allowing for a small variationof baselines around the ’centre-to-cen-tre’ baseline. Inside VINCI, optical fi-bres are used as spatial filters. The sizeof the optical fibres is exactly an Airydisk of the telescope, i.e. the field ofview is one Airy disk. This spatial filter-ing is equivalent to performing an aver-age over the visibilities inside the diskrepresented by the balloons.

For imaging it is important to note thatone measures as many independentimage points inside the Airy disk as onehas baselines, i.e. as one has degreesof freedom. Fortunately, the number ofbaselines is not only determined by thenumber of telescopes but also by theduration of the observation. Due to thesidereal motion, the baseline projectedonto the viewing direction changes allthe time in length and orientation de-pending on the object coordinates.Some of the baselines that are dis-played in Figure 2 have a slight peanut

19

Figure 1: The reconstructed, two-dimensional interferometric point spread function (PSF) ofthe star Achernar observed in the K-band. The width of the central fringe is 3 by 15 milli-arc-sec indicating the angular resolution limit of the VLTI for the baseline distribution shown inFigure 2. On the largest scale, the image is enveloped by the Airy disk of a single 8-m UnitTelescope. Its first minimum at 57 milli-arcsec off the centre can be clearly seen. The imageprovides a dramatic illustration of the 20-fold increase in resolution of the VLTI over a single8.2-m telescope.

shape due to the observing time of afew hours. Several individual meas-urements were taken during this time.Therefore, the number of individualpoints in the image will be considerablylarger than six when observing with allfour Unit Telescopes.

The present observations of Acher-nar give an impression of the pointspread function of the VLTI becauseAchernar with an angular diameter of 2milli-arcsec is smaller than the resolu-tion limit of 3 milli-arcsec of the VLTI inthe K-band. Due to the moderate num-ber of baselines, the point spread func-tion in Figure 1 shows a mix of a mod-erate number of individual fringe pat-terns. A larger number of visibilitymeasurements would make the individ-ual fringe patterns disappear, and onlythe central fringe would remain as the‘point’ in the point spread function.

The particular distribution of baselinesin Figure 2 is rather extended in thenorth-east/south-west direction (thebaseline of ANTU-YEPUN) and rathernarrow in the perpendicular directionbecause the baseline MELIPAL-YEPUNis missing. Consequently, the Fouriertransform of this distribution is rathernarrow in the north-east/south-west di-rection and rather wide in the perpendi-cular direction. The terms narrow andwide refer to the width of the centralfringe. The central fringe in Figure 1 hasa width of about 3 milli-arcsec in the di-rection of the longest baseline and ofabout 15 milli-arcsec in the perpendicu-lar direction. Due to the processingtechnique, the Airy disk of an individual

8-m telescope shows up in the pointspread function in Figure 1. The firstdark ring of the Airy disk with a radiusof 57 milli-arcsec in the K-band is read-ily apparent. This illustrates the gain inangular resolution with the VLTI com-pared to an individual Unit Telescope.

However, it is still a long way to go be-fore the VLTI will do imaging of arbitraryobjects since this requires measuringthe phase of the visibility function in ad-dition to its modulus i.e. its contrast.AMBER in 2003 will be a first step, be-ing able to use three telescopes andperforming a phase reconstruction viathe closure phase technique. Eventu-ally, from 2005 onwards, PRIMA will al-low measuring the phase directly perbaseline and, in addition, observe veryfaint objects. The scientific objectives ofthe PRIMA facility are described in [10].

3. OutlookThe coming year will be extremely

busy, with integrations of new subsys-tems about every two months. FINITO,the fringe sensor unit, will be first, fol-lowed by two MACAO systems, by thefirst two Auxiliary Telescopes, by AM-BER and by the carriages with cat’seyes of the three new Delay Lines. Thisis a big burden for the integration teamssince the VLTI is not the only ‘tele-scope’ at Paranal.

In 2004, the remaining two MACAOsystems and the third and fourthAuxiliary Telescopes will arrive atParanal so that the VLTI will be able tocombine the light from eight differenttelescopes, four Unit Telescopes and

four Auxiliary Telescopes. With sixDelay Lines, a common focus of six tel-escopes can be used at any given mo-ment. To fully exploit the VLTI infra-structure, the first second-generationinstrument should clearly be a six-waybeam combiner, measuring 15 base-lines at once and 28 during a night,without reconfiguring the telescopes.

Amongst all these integration andcommissioning activities one must notforget that the success of an observa-tory depends solely on the scientificoutput. MIDI will make a start next year,and 2004 will be largely devoted to sci-ence operations with the VLTI.

AcknowledgementOver the last 20 months, the number

of people both at ESO and in the com-munity contributing to the VLTI has be-come too large to name here. I wouldlike to thank not only those with thehighest visibility but also the large num-ber of contributors forming the base ofthe success by ensuring that every littlepiece of this complex machine worksevery night.

References

[1] Kervella, P., et al. 2002, in Exotic Starsas Challenges to Evolution, IAU Sym-posium 187, in press.

[2] Wittkowski, M. and Hummel, Ch. 2002, inInterferometry for Optical Astronomy II,ed. Traub, W., Proc. SPIE 4838, in press.

[3] Richichi, A. and Wittkowski, M. 2002, inThe VLTI: Challenges for the Future, eds.Garcia P.J.V., Glindemann A., HenningT., Malbet F., JENAM Workshop, in press.

20

▲ Figure 2: The set of baselines used for observingAchernar with the four Unit Telescopes. Each baseline isrepresented by two opposite, short arcs, symmetric aroundthe origin (centre) of the diagram. The colour-coded patternreflects the telescope pairs (ANTU-KUEYEN = magenta,ANTU-MELIPAL = red, ANTU-YEPUN = green, KUEYEN-MELIPAL = cyan, KUEYEN-YEPUN = blue).

Figure 3: The lines indicate the Unit Telescope and AT sta-tions that have been combined at Paranal to date. Thecolour scheme for the UTs corresponds to the baselinesshown in Figure 2. �

�N

[4] Segransan, D., et al. 2003, A&A Letters,in press.

[5] Paresce, F., et al. 2002a, in Inter-ferometry for Optical Astronomy II, ed.Traub, W., Proc. SPIE 4838, in press.

[6] Percheron, I., et al. 2002, in The VLTI:Challenges for the Future, eds. Garcia P.

J. V., Glindemann A., Henning T., MalbetF., JENAM Workshop, in press.

[7] Koehler, B., et al. 2002, The Messenger,this volume.

[8] Kern, P., et al. 2002, in Interferometry forOptical Astronomy II, ed. Traub, W.,Proc. SPIE 4838, in press.

[9] Arsenault, R., et al. 2002, in AdaptiveOptical System Technologies II, eds.Bonaccini, D., Wizinowich, P., Proc.SPIE 4839, in press.

[10] Paresce, F., et al. 2002b, in Interferom-etry for Optical Astronomy II, ed. Traub,W., Proc. SPIE 4838, in press.

21

The Auxiliary Telescopes for the VLTI: a Status ReportB. KOEHLER1, C. FLEBUS2

P. DIERICKX1, M. DIMMLER1, M. DUCHATEAU1, P. DUHOUX1, G. EHRENFELD3, E. GABRIEL2, P. GLOESENER2, V. HEINZ3, R. KARBAN1, M. KRAUS1, J.M. MORESMAU1,N. NINANE2, O. PIRNAY2, E. QUERTEMONT2, J. STRASSER, K. WIRENSTRAND1

1ESO, Garching, Germany; 2AMOS, Liège, Belgium; 3ESO, Paranal, Chile

1. Introduction

In June 1998, ESO signed a contractwith the company AMOS (Belgium) forthe supply of the Auxiliary TelescopeSystem (ATS) for the VLTI. The originalscope covered two movable AuxiliaryTelescopes (ATs), as shown in Figure 1,as well as their associated site equip-ment including rails and interface de-vices for each observing station. Anamendment was signed in September1999 for the supply of a third AT and,last September, a fourth AT was or-dered.

The contract with AMOS is based ontop-level performance requirementsand includes the design, manufacturingand testing of the complete system in-cluding all mirrors and cells, completemechanical structure, drives, encoders,small mechanisms and low-level elec-tronics. The main sub-contractors ofAMOS are FISBA (Switzerland) for thecoudé mirrors, PHASE (Italy) for themotors and CSEM (Switzerland) for theM2 hexapod mechanism.

ESO is in charge of the design anddevelopment of the telescope controlhardware and software, as well as ofthe two star sensors located at thecoudé focus.

After integration of the ESO controlhardware and software, the telescopeis fully tested in Europe at AMOS withthe possibility of sky observation pro-vided by an outside observing stationincluded in the dedicated test facility.After this, ESO transports the ATs toParanal, re-integrates them in the so-called Mirror Maintenance Building(MMB) and finally commissions themon the sky.

The project entered into manufactur-ing phase in mid-1999 and is nowreaching the end of a very extensivetesting and verification phase in Europeon AT#1 before its shipment to Paranalnext year.

This article recalls the rationale at theorigin of the ATS development, pro-vides a description of the design and fi-nally reports on the performance asmeasured in Europe so far.

2. Why Does the VLTI NeedAuxiliary Telescopes?

The VLTI is primarily intended tocombine coherently the four VLT 8-mUnit Telescopes (UTs). Obviously, theultimate VLTI sensitivity will indeed beobtained when combining the UTs.However, the arrayof 1.8-m diameterATs is a key ele-ment for the tech-nical and scientificcapability of the in-terferometer. Themain reasons arelisted below.

• It provides thebest imaging capa-bility of VLTI bycomplementing thearray of UTs. Itgives access to 30telescope stationsincreasing there-fore the number ofaccessible base-line vectors, a fun-damental parame-ter for high-fidelityimage reconstruc-tion.

• It gives accessto the longest VLTIbaseline of B = 200

m versus a maximum of 130 m be-tween UTs. This is needed to reach theultimate angular resolution of the VLTIthat scales as λ/B (i.e. 0.6 milli-arcsecin the visible and 2 milli-arcsec in the Kband).

• It enables full-time use of the VLTIfacilities, since the ATs are entirely ded-icated to interferometric observations.This is an important factor for the VLTIscientific productivity (and for the amor-tization of its development cost!).

• It is required by the NarrowAngle Astrometry mode of PRIMA that

Figure 1: The firstAuxiliary Telescopefor the VLTI duringfinal testing atAMOS in Liège(Belgium).