First Scientific Results with the VLT in Visitor and Service Modes

Home , NGC 6397, NGC 6544, NGC 6558, NGC 6717

No. 98 – December 1999 First Scientific Results with the VLT in Visitor and Service Modes Starting with this issue, The Messenger will regularly include scientific results obtained with the VLT. One of the results reported in this issue is a study of NGC 3603, the most massive visible H II region in the Galaxy, with VLT/ISAAC in the near-infrared Js, H, and Ks-bands and HST/WFPC2 at Hα and [N II] wavelengths. These VLT observations are the most sensitive near-infrared observations made to date of a dense starburst region, allowing one to investigate with unprecedented quality its low-mass stellar population. The sensitivity limit to stars detected in all three bands corresponds to 0.1 Mᓃ for a pre-main-sequence star of age 0.7 Myr. The observations clearly show that sub-solar-mass stars down to at least 0.1 Mᓃ do form in massive starbursts (from B. Brandl, W. Brandner, E.K. Grebel and H. Zinnecker, page 46).

Js versus Js–Ks colour-magnitude diagrams of NGC 3603. The left-hand panel contains all stars detected in all three wavebands in the entire field of view (3.4′×3.4′, or 6 pc × 6 pc); the centre panel shows the field stars at r > 75″ (2.25 pc) around the cluster, and the right-hand panel shows the cluster population within r < 33″ (1pc) with the field stars statistically subtracted. The dashed horizontal line (left-hand panel) indicates the detection limit of the previous most sensitive NIR study (Eisenhauer et al. 1998). The right-hand panel also shows the theoretical isochrones of pre-main-sequence stars of different ages from Palla & Stahler (1999) and the main sequence for dwarfs. For comparison, some corresponding stellar masses have been plotted next to the isochrones.

1 TELESCOPES AND INSTRUMENTATION The VLTI – The Observatory of the 21st Century A. GLINDEMANN, R. ABUTER, F. CARBOGNANI, F. DELPLANCKE, F. DERIE, A. GENNAI, P. GITTON, P. KERVELLA, B. KOEHLER, S. LÉVÊQUE, G. DE MARCHI, S. MENARDI, A. MICHEL, F. PARESCE, T. PHAN DUC, M. SCHÖLLER, M. TARENGHI and R. WILHELM

European Southern Observatory

1. Introduction servatory and with institutes and institu- smaller contracts for the Test Siderostats, tions from other European countries (Italy, for the Coudé Optical Trains of the four After several years on a bumpy road, Switzerland and Belgium) on the verge Unit Telescopes (UTs), and for a Feasi- the support for the VLT Interferometer of joining the project in one way or an- bility Study of the dual-feed facility PRI- project has increased dramatically over other. MA were also placed. the last year with the appointment of Major contracts for the Delay Line Sys- We are now planning for first fringes eight new staff members, with the foun- tems and for the Auxiliary Telescope Sys- with the commissioning instrument and dation of the NOVA1-ESO VLTI Expertise tems (ATs) were signed in 1997 and 1998 the Siderostats before the beginning of Centre (NEVEC) at the Leiden Ob- respectively, the work on the science in- the new Millenium. struments MIDI and AMBER has pro- In this article we will give a report on gressed very well in the last two years the status of the project and we will de- 1NOVA is the Netherlands Research School for and the commissioning instrument VIN- scribe the strategy for making the VLTI Astronomy. CI is coming closer to reality. Several the observatory of the 21st century.

Figure 1: The optical layout of the VLTI with two telescopes. The telescopes represent both UTs and ATs that have the same optical design. The Coudé Optical Trains are the mirrors after the tertiary mirror up to the Coudé Focus. Two Delay Lines are shown to demonstrate the principle of operation. The VLTI laboratory is represented by the beam-combining lens forming fringes.

2 2. The Sub-Systems of the VLTI

The layout of the VLTI is displayed in Figure 1, for the sake of simplicity with only two telescopes. A star at infinity il- luminates the apertures in the two telescopes with a plane wave that is guided through the Coudé Optical Trains into the Delay Line Tunnel. The delay in arrival of the light at telescope 1 with respect to telescope 2 is compensated by the delay lines so that the beams have zero Op- tical Path Difference (OPD) when they interfere on the detector in the VLTI laboratory. The field of view of the VLTI is 2 arcsec. However, the dual-feed facility PRIMA will allow picking two stars at the Coudé focus of ATs or UTs each in a 2 arcsec field of view and separated by up to 1 arcmin. The measurements of the contrast of the fringe pattern and of its phase, i.e. the position of the white light fringe with respect to the nominal zero OPD position are the tasks of the VLT Inter- ferometer. Doing these measurements Figure 2: The carriage for the cat’s-eye of the Delay Line System in the Clean Room at Fokker for many different baselines (different in Leiden. The carriage has three wheels running on the 65-m steel rails. The diameter of the in length and orientation) allows re- wheels is about 40 cm and the length of the carriage 2.25 m. constructing an image with the angular resolution of a single telescope with a diameter equal to the longest baseline. Since the longest baselines are 130 m for readily manufactured and they will be de- three telescopes and thus three baselines the Unit Telescopes and 200 m for the livered to Paranal in January 2000. The at the same time will allow the applica- Auxiliary Telescopes and since the spec- implementation of the Siderostat control tion of closure phase techniques elimi- ified precision for the measurements of software will be finished by May 2000. It nating the influence of atmospheric tur- the fringe contrast is very high, the re- is planned to have first fringes with the bulence on fringe position. Each AT will quirements e.g. for the Delay Line sys- Siderostats and VINCI in December be equipped with a tip-tilt system cor- tems are very tough. Here, the beam tilt 2000. recting for the fast image motion induced accuracy has to be better than 1.5 arc- by atmospheric turbulence. Under the seconds (15 milliarcsec on the sky) and Delay Line Systems seeing conditions at Paranal, tip-tilt cor- the absolute position accuracy is 1µm rection on a 1.8-m telescope in the near over 65 m of travel in the tunnel. Similar Three Delay Lines have been ordered; infrared means almost diffraction limited accuracy values are required for all mir- the first two will be installed in the De- image quality. One should note that the rors and mirror mounts of the VLTI, in- lay Line Tunnel in July 2000, the third one ATs are available exclusively for the VLTI, cluding the 8-m primary mirrors. Also, the in December 2000. The systems consist forming an observatory that is operated environmental conditions have to be very of the cat’s-eye mirror system retro-re- independently of the Unit Telescopes. stable in order to minimise internal tur- flecting the incoming parallel beam, and bulence. of the carriage transporting the cat’s-eye Coudé Optical Trains and Transfer It is the advantage of the VLTI that along the delay lines. Figure 2 shows the Optics these considerations have been driv- carriage on its three wheels in the cleaning the design of the Unit Telescopes room at the manufacturer. The three-mir- The five mirrors after the tertiary mir- and of the Auxiliary Telescopes, as ror optical design of the cat’s-eye has a ror of the Unit Telescopes form the Coudé well as the construction of the infra- variable curvature mirror (VCM, see [4]) Optical Train with a field of view of structure. Several measurement cam- in the focus of the cat’s eye in order to 2 arcmin (see Fig. 1). The last mir- paigns have confirmed that both opto- re-image the telescope pupil into a fixed ror before the Coudé Focus is very close mechanical and environmental specifi- position in the VLTI laboratory while the to the telescope exit pupil and can be re- cations are met. Delay Line System is tracking. The placed by the deformable mirror of the The status of the VLTI sub-systems is cat’s-eye can handle two input beams as adaptive optics system. The image qual- as follows: required for a dual feed system. A total ity at 2.2 µm is diffraction limited on of eight Delay Line Systems can be host- the optical axis and better than 97% over Test Siderostats ed in the Delay Line Tunnel. the full field of view. The Coudé Optical Trains for all four UTs have been or- For tests with the VLTI and the in- Auxiliary Telescope Systems dered and will be installed at Paranal in struments, two Test Siderostats were de- October 2000. signed and built (for details see [3]). The In June 1998, two 1.8-m Auxiliary Tele- The transfer optics between the Coudé free aperture of 400 mm allows for about scopes were ordered, and in June 1999, Focus and the Delay Lines are current- 20 stars to be observed in the N-band at the option for the third AT was exerted. ly being designed. The concept is to avoid 10 µm. Since for the VLTI Control Sys- A model of the telescope is displayed in the need of human intervention in the de- tem the Siderostats will ‘look’ the same Figure 3. The first two telescopes will be lay line tunnel when switching between as the UTs and ATs, it is planned to com- ready for the VLTI in June 2002, the third AT stations or UTs. Thus, the mirrors in mission the VLTI with the Siderostats. in October 2002. The telescopes are re- the Delay Line Tunnel reflecting the light They will be located on the AT stations locatable on 30 stations of the VISA (VLT into the Delay Lines and from the Delay and will thus be able to use baselines be- Interferometer Sub-Array) providing Lines into the VLTI Laboratory will be re- tween 8 and 200 m. The systems are baselines between 8 and 200 m. Using motely controlled.

3 provided by the ATs. The nominal position of equal optical path length is set for all beams at the same distance after the switchyard (indicated by the red line ZOPD in Fig. 4) to simplify the optical alignment of the interferometric instruments.

3. Phase A – Two Telescope Interferometry

In accordance with the recommendation by the Interferometry Science Advisory Committee (ISAC) laid out in the VLTI New Plan [12], the VLTI in its first phase will combine two telescopes, and near- and mid-infrared instrumentation will be available. An adaptive optics system for the UTs and a fringe sensor unit (FSU) will be implemented as well. In the near infrared, adaptive optics on the UTs is mandato- ry if the effective telescope aperture is not to be reduced to the size of the Fried Pa-

rameter, r0, which is about 1 m in the near infrared.

Figure 4: The layout of the VLTI Laboratory. The switchyard can direct the beam into four directions: (1) to the interferometric instruments Figure 3: A model of an Auxiliary Telescope during Observations. The 1.8-m telescope (with on the left without beam compression, (2) to an Alt-Az mount like the Unit Telescopes) is rigidly anchored to the ground by means of a spe- the interferometric instruments after beam cial interface. The light is directed via the Coudé Optical Train to the bottom of the Telescope. compression, (3) to the Differential Delay Lines From the Coudé Relay Optics it is sent on to the underground Delay Line Tunnel (DDLs) after beam compression, and (4) to the Differential Delay Lines without beam compression. The beam compressor reduces the beam diameter from 80 mm to 18 mm and it VLTI Laboratory for the interferometric instruments both re-images the telescope pupil into the PRIMA when observing with UTs or ATs. There- FSU, MIDI and VINCI (the position of the pupil The concept driver for the optical lay- fore, a beam compressor reduces the 80 is indicated by a diagonal line). A possible lo- out of the VLTI Laboratory (see Fig. 4) mm beam diameter from the UTs to cation for a visitor instrument is shown on the was to provide the same beam diameter 18 mm matching the diameter of 18 mm right. ĭ

4 VINCI – The Commissioning Instrument

The commissioning instrument of the VLTI is a conceptual copy of FLUOR, the near-infrared interferometric instrument of the IOTA interferometer on Mount Hop- kins in Arizona (US) [2]. The idea is to limit the technical risk and to facilitate the commissioning of the VLTI by using the design of an existing well functioning instrument (see e.g. [6]). The main component of VINCI (see Fig. 5) is a fibre beam combiner using the light from two telescopes as input and producing four outputs, two photometric and two interferometric signals. By varying the OPD between the beams with an internal modulator, a temporally modulated fringe pattern is produced on the detector. In addition to serving as the interferometric instrument, VINCI provides alignment tools and reference sources for the VLTI and the scientific instruments. VINCI will be delivered to Paranal in October 2000. Figure 5: The concept of the VLTI commissioning instrument VINCI. The beam combining unit, VINCI, and the reference source unit, LEONARDO, are placed on individual tables. Their respective positions in the VLTI Laboratory are shown in Figure 4. The main elements on the MIDI VINCI table are the fibre beam combiner and the dewar of the infrared detector. The inserted photographs show details of the FLUOR instrument at the IOTA interferometer. VINCI is a con- An interferometric instrument in the ceptual copy of FLUOR. mid-infrared at 10 µm in combination with 8-m telescopes is a novelty, promising exciting results on new types of ob- aging through phase closure techniques sensitivity is improved by using a bright jects. [10]. It is planned to start commissioning guide star for fringe tracking – simi- MIDI is being designed and built by a AMBER with the Siderostats in February lar to the guide star in adaptive optics for European consortium led by the Max- 2002. wavefront sensing – in one of the two Planck-Institute for Astronomy in Heidel- ‘feeds’ allowing increasing the expo- berg. The design philosophy is to have MACAO sure time on the science object in the oth- a simple instrument concept combining er feed up to 10–30 minutes depending two beams and providing a moderate The adaptive optics system MACAO on the position in the sky. With a spectral resolution (Ȃ 200). The chal- will have a 60-actuator bimorph mirror high-precision laser metrology system lenge lies in controlling the high thermal and a curvature wavefront sensor. The used to determine the internal optical background. It is estimated that the in- deformable mirror will replace one of path length, a dual feed system also pro- dividual beams will have an emissivity of the mirrors of the Coudé optical train, vides an imaging mode. The VLTI dual about 50%, which corresponds to an thus requiring no additional optical ele- feed facility PRIMA is described in the fol- equivalent photon noise on the sky of ments. As noted above, MACAO is es- lowing. 53 mJy per Airy disk (with λ/D = 0.26″ for sential for all near-infrared instrumenta- Imaging with an interferometer relies an aperture of 8 m) for a broad band 10 tion to be used with the Unit Telescopes on filling efficiently the UV plane with µm filter (∆λ = 4 µm). Since the signal- including the Fringe Sensor Unit operat- many different baseline vectors. Al- to-noise-ratio scales as S/N∝ D2 the avail- ing in the H-band. This means that also though this can be done by observing with ability of the Unit Telescopes is a huge a mid-infrared instrument like MIDI two telescopes in many different positions advantage over smaller apertures. The needs adaptive optics in order to improve using a dual feed system, it is more effi- details of the instrument are described in the limiting magnitude by using a Fringe cient to use three or more telescopes at [7]. MIDI will be delivered to Paranal in Tracker. the same time and eliminate the influence June 2001; first light with the Siderostats The first MACAO system will be in- of atmospheric turbulence by applying is planned for September 2001. stalled on one of the Unit Telescopes in phase closure techniques. Therefore, the December 2001. It is planned to have third and fourth delay lines as well as the AMBER MACAO ready for interferometric obser- third Auxiliary Telescope are included in vations with two UTs in June 2002. this phase. The near infrared instrument of the Clones of the MACAO system will be VLTI, AMBER, will operate between 1 and used for future VLT instrumentation like PRIMA 2.5 µm, at first with two telescopes with SINFONI. MACAO is an in-house de- a spectral resolution up to 10,000. As not- velopment [1]. The Phase Referenced Imaging and ed above, for UT observations in the Micro-arcsec Astrometry (PRIMA) facil- near-infrared adaptive optics is manda- 4. Phase B – Imaging ity is a dual-feed system adding a faint- tory. The magnitude limit of AMBER on object imaging and an astrometry mode the UTs is expected to reach K = 20 when The main drivers for the second to the VLTI [5, 11]. PRIMA enables si- a bright reference star is available (i.e. phase of the VLTI are increasing the sen- multaneous interferometric observations with a dual feed facility) and K = 14 oth- sitivity of the interferometer and adding of two objects – each with a maximum erwise. The European consortium in imaging modes. In addition, an astro- size of 2 arcsec – that are separated by charge of designing and manufacturing metric mode will open the door for many up to 1 arcmin, without requiring a large this instrument is led by the Universities thrilling scientific programmes. continuous field of view. One object will of Nice and Grenoble. AMBER has been Increasing the sensitivity of the VLTI then be used as a reference star for fringe designed for three beams to enable im- calls for a dual feed facility. Then, the tracking whilst the other object will be the

5 lutions and to obtain a thorough fi- nancial estimate for the manufacturing of the subsystems. The results of the study became available in July 1999 and a Call for Tender for the manufacturing of the system is planned early in 2000. Dual feed observations with PRIMA can start as soon as the star Figure 6: Principle of phase referenced imaging and astrometry with an interferometer. separator, the The difference in the positions of the white light fringe sensor unit fringes of object and reference star are de- and the differential termined by the OPD given by the product of delay lines are ∆S, the angular separation vector of the stars, ready. Then, a ref- and B, the baseline vector, by the phase φ of erence star can be the visibility function of the science object, by used for fringe the OPD caused by the turbulence, and by the tracking while inte- internal OPD. grating on the fainter science object as described science target. As a detector for PRIMA above. If the fringe either the two scientific instruments pattern can be sta- MIDI and AMBER can be used making bilised over 10– use of the fringe stabilisation provided by 100 sec the ex- PRIMA, or a dedicated PRIMA detector pected limiting Figure 7: Functional block diagram of the VLTI and the PRIMA com- for high-precision astrometry that will be magnitudes are ponents. The VLTI has the following subsystems: Telescope 1 and 2 designed in the next years. about K Ȃ 16 and – two UTs or two ATs, AO – Adaptive Optics, Delay Lines, AMBER/MIDI PRIMA is the key to access: (1) High- N Ȃ 8. – Science Instruments. The PRIMA subsystems are: Star Separator, er sensitivity, the limiting magnitude will Phase informa- Differential Delay Line, Laser Metrology, Fringe Sensor Unit. be about K = 20, (2) imaging of faint tion required for objects with high angular resolution imaging and as- (<10 milliarcsec), (3) high-precision astrometry becomes available if the laser astrometry. PRIMA shall be operational trometry (Ȃ 10 µarcsec over a 10-arcsec metrology system is installed. An OPD by 2003. field). measurement accuracy of 500 nm rms The principle of operation relies on over 10 min sets the limiting magnitudes 5. First Fringes and Beyond finding within the isoplanatic angle (Ȃ 1 to about K Ȃ 20 and N Ȃ 11. The Strehl The major challenge in making the arcmin) of the science target a sufficiently ratio in the reconstructed image can be VLTI work is the complexity of the sys- bright star (H Ȃ 12) that can be used as as good as 30% in the K-band and tem. Not only are the individual sub-sys- a reference star for the stabilisation of the 80% in the N-band depending on the tems high-tech instruments but they are fringe motion induced by atmospheric UV coverage. Reaching the final goal of also scattered over an area up to 200 m turbulence (see Fig. 6). Controlling all op- 5 nm rms over 30 min allows 10 µarcsec in diameter. The quality of the control sys- tical path lengths of the reference star and of the science star inside the interferometer (OPDint) with a laser metrology system introduces the capability of imaging faint objects and of determining the precise angular separation between the two stars. The measurement has to be repeated for up to 30 min in order to average out the variations of the differential OPD caused by atmospheric turbulence (OPDturb). PRIMA can be subdivided into the four sub-systems (see Fig. 7): Star Sep- arator, Laser Metrology System, Differ- ential Delay Lines and Fringe Sensor Unit. These sub-systems contain a number of technological challenges, in particular in the area of laser metrology, that require careful analysis of the possible technical solutions. Therefore, Figure 8: The UV coverage on the left and the point spread function (PSF) on the right with a a feasibility study for the four subsystems full width at half maximum of 4 mas respectively 8 mas in the narrow respectively wide direc- was performed by Dornier Satellite Sys- tion of the PSF at 2.2 µm. The UV coverage and the PSF are calculated for –15° declination tems, Friedrichshafen, and by ONERA, and 8 hours of observing when combining all four UTs. Producing images with this quality is Paris, to find the best technical so- the ultimate goal for the VLTI.

6 BER at the end of results. The full scientific potential of the 2001. Both science VLTI in its different phases is described instruments will be in [9] summarised by F. Paresce [8]. commissioned with the Siderostats. 6. The Future MIDI can be tested shortly after first The scenario of VLTI observations de- fringes on the scribed in the article will without any doubt Siderostats with provide a wealth of scientific discoveries the UTs even be- but it still does not exploit the full capac- fore the adap- ity of the infrastructure at Paranal. The tive optics system layout of the interferometric tunnel and MACAO arrives. of the VISA array allows combining more After installation than three telescopes providing either an and commission- even better coverage of the UV plane or ing of MACAO with the operation of different instruments at the VLTI and with the same time. VINCI in the sec- Therefore, the four final Delay Line ond half of 2002, Systems and all remaining Auxiliary Tele- AMBER will be scopes for a total of 8 ATs shall be in- tested on the UTs. stalled for VLTI operations. The comple- Also in 2002, a tion of the final phase will allow VLTI ob- total of three Aux- servations with 28 simultaneous base- iliary Telescopes lines. The planning also has to include will be commis- new, the second-generation, instrumen- sioned, first indi- tation allowing the combination of six to vidually then with eight telescopes and at the same time the VLTI and VIN- providing a dual feed facility for each pair CI. The science in- of telescopes. struments will fol- Beyond the fully equipped VLTI there low suit. The first have to be kilometric arrays with 8+ m tel- sub-systems of escopes in order to drive the angular res- PRIMA arrive as olution into the sub-milliarcsec range and Figure 9: The type of results obtained with VINCI. The two top rows well in 2002. Dur- to improve the sensitivity well beyond K display the photometric signals from each telescope, the third row dis- ing the course of = 20. The VLTI will be the stepping stone plays one of the interferometric outputs and the fourth row displays the 2003 both dual towards these observatories. same signal as in the third row calibrated with the photometric signals feed and closure to obtain a clean fringe pattern. phase operation References can be tested with the VLTI. [1] Bonaccini, D., Rigaut, F., Glindemann, A., tem is a key issue for the smooth oper- In order to operate the VLTI as an ob- Dudziak, G., Mariotti, J.-M. and Paresce, ation of the complete system. From the servatory, support astronomers, instru- F. 1998, Proc. SPIE 3353, 224–232. beginning of 2000 onwards, every 6 ment operators and engineers are re- [2] Coudé du Foresto, V., Perrin, G., Mari- months major new sub-systems and in- quired for regular operations and for otti, J.-M., Lacasse, M. and Traub, W.A. struments arrive at Paranal and have to 1996, Integrated Optics For Astronomi- maintenance. Although the number of cal lnterferometry, Grenoble, P. Kern and be integrated into the VLTI. Thus, the nights with two (or more) UTs is proba- F. Malbet eds., 110–125. strategy for commissioning has to be de- bly restricted to a small fraction of the to- [3] Derie, F., Brunetto, E. and Ferrari, M. fined very carefully. tal number of available nights in the first 1999, The Messenger 97, 12–13. For the first milestone of the VLTI pro- few years, first the Siderostats and then [4] Ferrari, M. and Derie, F. 1999, The Mes- gramme, First Fringes with the Sidero- the Auxiliary Telescopes are available senger 97, 11. stats and VINCI, we have adopted the fol- every night for VLTI observations. Thus, [5] Glindemann, A. and Lévêque, S. 1999 lowing approach: after installation and the planning for the VLTI has to foresee VLT Opening Symposium, in press. commissioning of the Siderostats, of the a full-time coverage with technical and [6] Kervella, P., Coudé du Foresto, V., Delay Lines and of the Transfer Optics, scientific personnel. One has to keep in Traub, W. A. and Lacasse, M.G.1999, ASP Conf. Series XX, 51. the optical alignment of the full system will mind that the VLTI is running in parallel [7] Leinert, Ch. and Graser, U. 1998, Proc. be done with a Technical CCD. Then, all to the VLT observatory with Unit Tele- SPIE 3350, 389–393. degrees of freedom but one can be ad- scopes and requires the same level of [8] Paresce, F., et al. 1996, The Messenger justed and the dynamic behaviour of the technical support for telescopes and in- 83, 14–21. delay lines can be tested with a star ob- struments. [9] Paresce, F., Ed. 1997 Science with the served with the siderostats. With the The final goal of the VLTI is to pro- VLT lnterferometer. Springer-Verlag. commissioning instrument VINCI the last duce images with a few milliarcseconds [10] Petrov, R., Malbet, F., Richichi, A. and Hof- remaining degree of freedom, the optical resolution. Figure 8 shows a simulated mann, K.-H. 1998, The Messenger 92, path difference, will be adjusted to pro- point spread function when observing 11–14. duce interferometric fringes before the for eight hours with all four UTs. The re- [11] Quirrenbach, A., Coudé du Foresto, V., end of 2000. After first fringes, in the first sult shows a very impressive albeit elon- Daigne, G., Hofmann, K.-H., Hofmann, R., Lattanzi, M., Osterbart, R., Le Poole, half of 2001, the VLTI will be optimised gated PSF with a full width at half max- × R., Queloz, D., Vakili, F. 1998, Proc. SPIE by using different AT stations for the imum (FWHM) of about 4 8 milli- arc- 3350, 807–817. Siderostats and by integrating the third sec for a wavelength of 2.2 µm. Having [12] von der Lühe, O., Bonaccini, D., Derie, Delay Line into the system. While this is this goal in mind, one must not forget that F., Koehler, B., Lévêque, S., Manil, E., taking place the Coudé Optical Trains in for several years the results will be in- Michel, A. Verola, M. 1997, The Mes- the UTs and the remaining transfer op- dividual measurement points in the UV senger 87, 8-14. tics are being installed. plane represented by curves like those The next instruments to arrive at in Figure 9. However, this does not di- Paranal are MIDI in the middle and AM- minish the scientific content of these E-mail: [email protected]

7 Laser Guide Star Facility for the ESO VLT D. BONACCINI, W. HACKENBERG, M. CULLUM, M. QUATTRI, E. BRUNETTO, J. QUENTIN, F. KOCH, E. ALLAERT, A. VAN KESTEREN

European Southern Observatory

1. Introduction and very low read noise are required. seeing at 0.5 µm, 6 msec τo, and 0.4 Thus photon noise is the only unavoid- Strehl contribution from field anisopla- ESO has been studying the possibil- able contributor to the WFS noise. Even natism at 20" elongation, in K-band. ity of introducing a Laser Guide Star so, with Natural Guide Stars (NGS) the The analytical simulations allow a Facility (LGSF) for the VLT to serve the performance in the K-band of a medi- parametric exploration of the AO per- adaptive optics (AO) instruments fore- um-sized AO system drops significantly formance. Although this is less accurate seen on the UT3 telescope of the when mV ≥15. Figure 2 shows the ex- than a full numerical simulation, the re- Paranal Observatory: NAOS-CONICA pected K-band Strehl performance for a sults are according to our estimates ac- and SINFONI. 60-element curvature system with the curate to within 10-15%, and serve well Although full project funding, accord- VLT as a function of NGS magnitude the purpose of this discussion. The ef- ing to the ESO Long Range plan, will and its angular distance from the sci- fect of anisoplanatism on AO systems is only start in 2001, we have performed a ence target. Note that Strehl curves de- very evident in Figure 2. For example, a comprehensive feasibility study, with pend by the seeing at 6th power. Our 10th magnitude NGS which is 25 arc- some innovative developments, espe- analytical computations are for 0.66" seconds away from the science target cially on the possible injection of the laser light through a single-mode fibre (see the following article). The LGSF concept is the basis of this article. The LGSF is foreseen as an add-on to the VLT telescope, able to create an artificial guide star in the mesospheric sodium layer, at 90 km average altitude. The LGSF would be used as a slave system to the AO systems, to produce an artificial reference star for the wavefront sensor, close to the observed object, whenever necessary. Figure 1 il- lustrates how the baseline LGSF concept would be implemented on the VLT. The activities reported here have been carried out at ESO, with the enthusiastic scientific collaboration of the Max-Planck-Institut für Extraterrestri- sche Physik (MPE) in Garching,. MPE has built the Calar Alto Laser for the ALFA system (Davies et al, 1999). The definition of the LGSF is at the conceptual design stage, with a solid baseline solution explored in detail and two open options: the final choice for the laser, and the use of a fibre relay for the laser beam instead of a mirror system. The final choices will be made during the LGSF Conceptual Design Review. This article gives an overview of the conceptual design we are currently working on.

2. Project Rationale

The most important limitation of current NGS-AO systems, is the scarcity of a good reference star for the wavefront sensor (WFS) to exploit the full potential gain of the adaptive optics loop. The NGS should be bright enough to overcome both sensor read-out and photon noise. If it lies outside of a small field, optimum correction cannot be achieved, due to anisoplanetism effects. Figure 1: Layout of the LGSF. The laser and its controls are mounted under the Nasmyth plat- To minimise the former, wavefront form, in a well-insulated laser room. The laser beam is relayed to the 50-cm-diameter launch sensors with a high quantum efficiency telescope located behind the UT secondary mirror.

8 Figure 2: K-band iso-Strehl curves, computed for a 60-element cur- Figure 3: Logarithmic iso-probability curves for NGS. V-band NGS mag- vature system, vs NGS magnitude (x-axis) and elongation between nitude on the horizontal axis, NGS elongation from the science object science object and NGS (y-axis, arcsec). in arcsec, on the vertical axis.

gives the same Strehl ratio as a 17th made equivalent to that of a NGS of mV tion to the LGS. The image blur is cor- magnitude NGS on-axis. ≅ 9.5 with seeing-limited diameter. It is rected using the LGS, the image motion The scientific justification of the rather like providing the telescope with (tip-tilt) using a NGS. The NGS for tip- LGSF comes from sky-coverage con- a headlight, enabling the AO system to tilt can, however, be much fainter and siderations. Put in simple terms, we find a reference star wherever the tele- typically 2.5 times further away than a need the answer to the question: ‘what scope points. This opens the possibility high-order NGS reference. fraction of the sky can be observed with of observing many more extragalactic The ultimate sky coverage and AO AO, and how many of my objects can targets than with a NGS-AO system. performance depends largely on the be observed?’ (Le Louarn et al., 1998, Two problems still prevent us reaching tip-tilt sensor and is one of the reasons Di Serego Alighieri et al., 1994) The re- complete (100%) sky coverage: (a) the why ESO has put considerable effort quirement of a relatively bright NGS, at ‘focus anisoplanatism’ or cone effect to develop a high-performance tip-tilt short distance from the science object, (Tyler, 1994), and (b) the intrinsic inde- correction subsystem, the so-called limits the use of AO systems. The high- termination of the image motion (tip- STRAP system. We will report on the er the Strehl Ratio required the greater tilt) signal. The cone effect, for an 8-m STRAP system in a next issue of The the limitation. Figure 3 shows the medi- telescope in the Near Infra-Red (NIR), Messenger. an sky coverage iso-probability curves and one single LGS, gives an unavoid- The K-band iso-Strehl contours for a for a NGS plotted on a logarithmic able but tolerable reduction of the LGS-AO 60-element curvature sensing scale. Strehl performance. system are shown in Figure 4. The Using a LGS, the reference signal The tilt problem can be overcome by NGS magnitude and elongation axis re- for the AO wavefront sensor can be using a natural star in the field in addi- fer to the tip-tilt reference star. The LGS is assumed to be pointed close to the science target. Compared to Figure 2, Figure 4: K-band iso-Strehl curves, the gain in sky coverage is very evident for a 60-element even though the cone effect limits the curvature sensing maximum achievable Strehl. There is AO system with a an even larger gain if one looks at the LGS, using a tip-tilt encircled energy improvement for a natural guide star. spectrograph (Le Louarn et al., 1998, The NGS V-magni- Bonaccini, 1996). tude and elongation Figure 5 combines the information (arcsec) from the contained in Figures 2 and 4, and science target are the horizontal and shows the overall sky coverage (com- vertical axes re- bining NGS magnitude and elongation) spectively. This plot achieved with a given Strehl ratio, for should be compared NGS-AO and LGS-AO systems. The with Figure 2. simulated system corrects 80 Karhu- nen-Loewe modes1. Using a LGS-AO,

1Karhunen-Loewe modes: they are an orto-normal basis for the expansion of the wavefront. They are often used in AO systems, as the modes are also statistically independent (see e.g. Adaptive Optics in Astronomy, p. 31, F. Roddier ed., Cambridge Press University 1999).

9 pare a proposal for equipping with it the VLT Unit Telescope hosting AO systems. A baseline solution has been studied in detail during 1998, for the subsystems of the laser, the beam relay and the launch telescope. The preliminary design activities were frozen in February 1999, except for the studies on a single-mode fibre relay. This was to allow the feasibility of the fibre relay system to be thoroughly investigated before proceeding further. The past months have been spent on the development of a prototype single- mode fibre relay system, from the laser output to the launch telescope. The fibre-optics industry has been involved in order to have access to state-of-the-art fibre production technologies. We have used a 532 nm, 5W Continuous Wave (CW), Verdi* laser from Coherent GmbH, to do the laboratory experiments and studies. The results of our Figure 5: Sky coverage of an AO system correcting 80 Karhunen-Loewe modes, as a function studies and tests so far indicate that the of the K-band Strehl Ratio. The model atmosphere is as described in the article. NGS and LGS fibre relay solution, although not as curves are plotted. The assumed single sodium LGS has a mV = 10 and 1 arcsec FHWM. The easy as it seems, is feasible and also saturation of the LGS curve at SR = 0.52 is due to the cone effect: Irrespective of how good very attractive. We plan to conclude this the tip-tilt reference signal is, higher Strehl ratios are not possible. development with working fibre relays in June 2000, and resume LGSF de- the maximum K-band Strehl obtainable and for the ‘perfect’ telescope. The tailed design activities in the summer is 0.52 in our example, and this is seen LGS-AO system allows the attenuation 2000. Assuming that funding is forth- in the saturation of the LGS-AO sky of all the spatial frequencies transmitted coming as planned, one could envisage coverage curve. The sky coverage for a by the telescope to be restored by post- commissioning of the LGSF on the VLT Strehl Ratio < 0.45 in K-band is always facto deconvolution. by the end of 2003. larger than NGS-AO in this configura- Techniques to reduce or remove the A brief explanation of the baseline tion. LGS cone-effect are being studied at LGSF design is given below and repre- The LGS-AO system foreseen for the various institutes, for example by using sents the present conceptual design VLT will not reach Strehl ratios much multiple guide stars (Fried and Belsher, around which we are working. It is like- greater than 40% in K-band, due to the 1994), or with a hybrid system where ly to be modified somewhat during the cone effect. Note that the PSF core the low orders are corrected with the preliminary and final design phases. achieves the resolution of the diffraction NGS, the remaining high orders with limit (0.057″ in K-band), therefore the the LGS. These will be necessary for 4. High Level Operational and spatial information can be recovered. visible AO compensation on future ex- Functional Requirements The images may be further boosted tremely large aperture telescopes. using post-processing techniques such The LGSF is seen by the AO instru- as deconvolution if necessary (Conan 3. Project Status ments as a telescope server facility. It is et al. 1998, Christou et al., 1999). Fig- therefore compliant with the VLT re- ure 6 shows the long-exposure Modu- ESO’s Scientific Technical Commit- quirements/standards and integrated lation Transfer Function (MTF) for the tee has confirmed the scientific rele- hardware and software with the tele- uncorrected seeing, for the LGS-AO vance of the LGSF, and asked to pre- scope environment. The AO systems (NAOS and MACAO) on UT3 will be able to send basic commands and re- ceive the diagnostic status. The LGSF will never take independent initiatives except for safety reasons, such as air-traffic alerts from the dedicated monitoring camera. In this case, it notifies to the clients that it is going to shut down the beam propagation before doing it. Concerning light pollution from the LGS, work is in progress to determine it quantitatively (Delplancke et al., 1998). The monochromatic laser beam scattering does not disturb the IR instruments. The operational scheme proposed is to create an area-of-avoidance around the laser beam: the LGSF monitors the pointing of the other Paranal telescopes, and creates an alert status

Figure 6: MTF in K-band, for the VLT with uncorrected seeing, LGS-AO correction, and the *The Verdi is an all-solid-state diode pumped perfect telescope Nd:YVO4 laser, recently produced by Coherent.

10 Figure 7: Baseline ed behind M2. The aperture is dictated laser conceptual by the optimisation of the LGS beam di- scheme. Two solid- ameter for the best Paranal seeing state pump lasers conditions. The projected beam diame- are combined in a ter will be adjusted to the actual seeing single beam, 20W at 532 nm, pump- conditions. The telescope is a polychro- ing a ring-dye matic reflective design, which allows resonator laser, the LGS to be positioned within a 1-ar- using Rodhamine cminute field on the sky, with a visible 6G as dye and monitoring camera. Figure 1 shows a servo-tuned to schematic diagram of the foreseen set- 589 nm. up. The level of automatism to be built into the LGSF is large, in order to allow routine and reliable operation at the VLT. This increases the initial cost and complexity of the facility, but we believe it will prove highly beneficial later, during operation. A number of servo-systems and automatic computations are when another telescope is pointing the laser room to the launch telescope planned, based on dedicated hard- close to the cone-of-avoidance. The ac- could in principle use flat mirrors in a ware/software. tions to take in this case are still to be shielded light-path. The laser beam decided. would be hidden behind one of the 6. Baseline Laser and Laser The LGSF is required to produce a secondary mirror (M2) support spiders Options mV = 9.5 equivalent artificial reference when crossing above the primary mir- star with a FWHM of 1.5″ (seeing limit- ror. The mirror relay system was in- Only a complex micro-macro pulse ed as goal). For median Paranal see- vestigated in detail. It is cumbersome modulation scheme would allow slightly ing, down to 60 degrees zenith dis- and difficult to implement in practice. greater flux returns from the mesos- tance, a 6.5W CW Sodium laser will Turbulence effects in the relay path pheric sodium than a CW laser (Milonni serve the purpose. For worse seeing as well as dust contamination and and Fugate 1997). Considering that conditions, or in the presence of thin cir- thermal effects on the optics, can de- CW lasers are intrinsically more stable rus clouds, a laser power in the range grade the LGS quality considerably. and simpler, that coatings for CW lasers of 10W would be preferable. Air tightness, sev- The LGS has to be independently eral servo-con- pointed within 20″ from the defined op- trols and also tical axis of the VLT, with a relative frequent mainte- pointing accuracy of 0.3″. Automatic fo- nance would be cussing of the LGS on the sodium layer necessary. will be performed, up to zenith angles The alternative of 60 degrees. solution for the The safety measures of the LGSF laser relay we will comply with the Paranal observato- propose uses a ry strict safety regulations, Chilean law, 30 m long, single- and the FAA regulations as applicable mode optical fi- in Chile. bre. This is an innovative solution and not easy, 5. LGSF Baseline Design given the power Solution densities involved: one Watt of inject- The baseline solution study has di- ed power in a 5.5 vided the LGSF into three different sub- micron core fibre systems: (a) the laser room, including corresponds to a the laser and its servo-controls, (b) the power density of beam relay and (c) the launch tele- 4 MW/cm2. We scope with its diagnostics. We report have found a vi- here briefly on (a) and (c). A companion able solution to article in this issue describes the proto- this problem. It type fibre beam relay subsystem in de- will deliver a dif- tail. We will report in future issues of fraction-limited The Messenger design details on sub- beam at the systems (a) and (c). launch telescope We propose to use a relatively low- input, with >70% consumption laser (4KW), and to build throughput. The a thermally stabilised laser room un- use of a single- der the Nasmyth platform of UT3. The mode fibre would laser choice takes into account com- result in savings mercial availability, support, stability in laser power, and servicing issues, so as to minimise money and man- the LGSF operation and overhead power. costs. A 50-cm diam- Figure 8: Layout of the laser room designed, with the laser optical table The baseline solution for the relay eter laser launch and the necessary accessories. The room is thermally isolated and con- of the 20-mm-diameter laser beam from telescope is locat- ditioned.

11 are less demanding, and that CW vices. The laser Germanosilicate lasers safety issues are easier to han- optical table size Yb: Double-Clad Fibre Fibre dle than for pulsed lasers, we are con- is chosen so as to centrating on CW lasers solutions. allow a later addi- The baseline laser is a modified ver- tion of a second Diode sion of a commercial product from Co- 6.5W laser. The herent, made of two 10W Nd-YV04 CW room is thermally Verdi diode laser pumps at 532 nm, com- isolated and con- bined into the Coherent jet-streamed ring ditioned. dye resonator model 899-21, using Rho- The total mass damine 6G as dye. The general advan- of the laser bench tages of solid state lasers as compared equipment is es- Figure 9: Fibre laser layout. The pump laser is a diode-pumped dou- to ion-lasers for pumping, are the much timated to be 1.3 ble-clad Yb-doped fibre that outputs 1.1137 µm. The Raman is a high- higher conversion efficiency and greater metric tons. We delta germanosilicate single-mode fibre, which is directly coupled to compactness. estimate the en- the pump fibre. Feedback for both the pump and Raman fibres are pro- In addition the wavelength of fre- tire laser room to vided by FBGs. The second harmonic of the 1.178 µm first Stokes is quency-doubled light from Nd-doped weight not more generated within the fibre Raman laser cavity, and is ejected by a dichroic laser crystals is around the maximum than 4 tons. An mirror. (Courtesy J. Murray, LiteCycles Inc.) of the absorption spectrum of Rh6G, analysis of the whereas the light of argon-ion lasers is static and dynamic impact of the laser 7. Alternative Laser Choices clearly off the absorption peak. room mounted under the Nasmyth plat- This laser configuration delivers > form on UT3 has been done and found Two alternative laser choices are 6.5W of CW power on a 20 MHz to be negligible. The power, communi- currently being monitored. FWHM TEM00* line, or a LGS with mV cation and cooling budgets have indi- One solution, in which the fibre itself ≅ 9 after beam conditioning. A similar cated that one VLT Service Connection is the laser, is being studied by Lite- ring-dye laser has proven to be robust Point (SCP) will be sufficient for the cycles Inc. (Murray et al., 1998). This is and stable during the past years at laser room. still in the feasibility phase and it has Calar Alto. The laser line is broadened to 500 MHz, to avoid both the saturation of mesospheric sodium atoms and non linear effects in the fibre. We have the capability of doubling this power in the future, by adding two such laser systems on the same optical table. This scenario would produce 13W CW of sodium laser power, and would use two independent single- mode fibres to relay the beams to the launch telescope. The orthogonally polarised beams at the fibre output would be combined in the launch telescope. For safety and operational stability reasons, the laser bench would be installed in an isolated room underneath the Nasmyth platform. This room also contains the dye solution chiller, the pump laser heat-exchanger, two electronics cabinets and workspace for maintenance work (Fig. 8). The dimensions of the laser room allow convenient access to the optical table. The room is temperature stabilised and supplied with filtered air to avoid contamination of the optics. In order not to impact on the telescope seeing the laser room is thermally insulated, with the outer walls at the same temperature as the air in the telescope enclosure. Vibration, air turbulence, temperature variations and dust contamination directly influence the line width as well as the power and beam-pointing stability of a jet-streamed dye laser. The dye-laser os- cillator therefore is installed in a fixed spatial position on a rigid optical table that de-couples the laser head from vibra- tions, e.g. when the telescope is moving. Figure 8 shows the layout of the laser room together with the accessory de- Figure 10: The launch telescope is a 25X beam expander. It is mounted behind the M2 unit, together with its accessory beam monitors: power, alignment, polarization meters, and beam pointing/focussing servos. Two cameras are present, one to monitor the LGS in the field, and *TEM00 is the fundamental Transversal Electro- magnetic Mode of the resonator. one for air-traffic surveillance.

12 recently attracted the interest of the LGS). The launch Gemini AO project as well. The con- telescope is thus a cept is shown in Figure 9. 25X beam ex- Two 100W CW laser diode-bar arrays pander. We have are used, efficiently coupled in a double- explored a reflec- clad Yb-doped fibre which outputs >50% tive solution based of the input power at 1.1137 µm from the on a two-parabolic- single-mode fibre core. The Yb-doped fi- mirror stigmatic de- bre is fusion-spliced with a single-mode sign and verified germanosilicate fibre 1 km long. The the feasibility of its Stimulated Raman Scattering (SRS) ef- opto-mechanical fect in this long fibre is used to stretch the tolerances. Light- wavelength and tune it to 1.178 µm, i.e. weight mirrors can twice the 0.589 µm of the D2 sodium line, be used; the total by the use of fibre Bragg gratings (FBG) weight of the LT written on the fibre and servo-tuned by and its structure is piezos. The fibre output beam is focused estimated to be on a second harmonic generation crys- 80 kg. tal (SHG) for intracavity doubling, which A wedged, anti- will convert the 1.178 µm photons to reflection coated 0.589 µm. The longitudinal mode spac- exit window pro- ing of such a long fibre Raman laser vides shielding from is 100 kHz, giving a truly broad-band ex- turbulence and dust, citation of the mesospheric sodium. The and a 10–3 laser goal for the line FWHM is ≤ 2 GHz, flux return, to mon- and the CW output power should be in itor the exit beam excess of 10W CW, with a goal of 30W. quality and power. However, an unpolarised sodium 2 GHz We will also moni- laser line would return about half the flux tor the beam polar- of a 20 MHz FWHM line. A 10W broad- isation state and its band laser would therefore be equiva- frequency spec- lent to 5W of ‘narrow-band’ power. Per- trum. The volume Figure 11: View of the laboratory setup where the fibre experiments ceived risks of this technology are the from the fibre end have been carried out. The setup allows the measurement of the spec- larger than expected linewidth, excessive to the exit window trum, power and propagation properties of the laser beam. The in- non-linearity (Stimulated Brillouin Scat- is airtight. jection, fibre output and SBS return beams can also be characterised. tering) losses and a lower than expect- The space avail- ed efficiency of the intracavity SHG crys- able behind the UT3 secondary mirror The pointing and steering would be tal. We are taking advantage of our mod- is limited and we are forced to expand done at the level of the fibre output. The els of the non-linear effects, developed the beam by a factor 25 within 600 mm range of fibre motion is ± 0.79 µm for for the fiber relay studies, to further in- of design space. The current design is slow pointing with 0.1 µm precision, and vestigate this laser design. shown in Figure 10. It delivers a very ± 12 µm for fast steering. Commercial The great advantages of this solution good performance with a central obscu- devices are available which meet these are that the single-mode fibre output ration of 0.7% in area for a 2-ar- specifications together with their control plus frequency doubling crystal give a cminute-diameter field (see Table 1). electronics. high beam quality, the long fibre length However it has tight tolerances and re- We plan to enclose the launch tele- (1 km!) allows the laser source to be quires an athermalised design: its fea- scope and its liquid-cooled electronics placed far away from the telescope, in sibility has been checked. We will also in an aerodynamic wind-shield, behind a VME-rack sized box. If this laser have to introduce alignment diagnos- M2 unit. The static and dynamic impact would be commercially available, it tics. The pointing has to be remotely on M2 structural stability have been would be the ideal solution. controlled within a minimum range of 1 studied and shown to be negligible. The second laser choice under in- arcmin radius. Fast steering is also nec- vestigation is a recent development at essary to compensate the LGS jitter in- MPE in Garching. It is a two-stage laser duced by the atmosphere on the up- Conclusion with a seeded dye-laser amplifier which ward-going 50 cm beam which differs would deliver at least 10 W CW with 500 from the downward path toward the col- The conceptual design of the LGSF MHz line width. This laser would give an lecting 8-m telescope. The expected has been completed and the main in- equivalent LGS magnitude mV = 8. rms jitter is 0.34 arcsec for the median terfaces with UT3 as well as the envi- Paranal atmosphere and therefore a ronmental and structural impacts have 8. Launch Telescope fast steering range of ± 2-arcsec is been assessed. Further optimisation specified. The steering signal is provid- will be carried out in the next phases. The input interface to the launch tel- ed by the AO high order WFS that is Laser beam relay system designs using escope (LT) is a collimated beam 20- looking at the LGS. The bandwidth of both conventional optics and mono- mm in diameter. The output is a diffrac- the steering device has to be better mode fibre have been explored. This tion limited spot at 90 km altitude with than 200Hz at –3dB. latter solution has many advantages the possibility to focus it between 70 and 200 km. The LT optical quality specification is to have a Strehl ratio Table 1: Launch Telescope Design – Optical Performance >0.95 within 10 arcsec of field radius, and with 80% encircled energy diame- Design Requirements / Value Achieved ter of 0.7" within 1 arcminute field radius. This is required to ease the align- Field 10 arcsec radius 1 arcmin radius ment tolerances and to have a CCD Strehl Ratio at 589 nm >0.95 / 0.99 -- / 0.975 monitoring camera looking at the sur- 80% Encircled energy diam. (“) -- / 0.45 0.7 / 0.5 rounding sky field (notching out the

13 and we now feel confident that this so- tive Optics Images with imprecise knowl- AO on 8m telescopes’, SPIE Proceedings lution can be adopted. edge of the point spread function: results Vol. 3353, 364, 1998. We have however to finish the on astronomical objects’, in ESO/OSA Top- Milonni, P. W. and Fugate, R.Q.: ‘Analysis of tests of the final fibre-relay modules in ical meeting on Astronomy with Adaptive measured photon returns from Sodium the laboratory. A burn-in test is being Optics: Present Results and Future Pro- Guide Stars’, in Proceedings of the ESO grams, p. 121, D.Bonaccini ed., 1998. Workshop on Laser Technology for Laser discussed with MPE, to be performed Davies,R, Hackenberg, W., Eckart, A., Ott, T., Guide Star Adaptive Optics Astronomy, N. with the 4.5W CW sodium laser at Butler, D., and Casper, M: ‘The ALFA Laser Hubin ed., p. 77, 1997. Calar Alto. After these tests have been Guide Star operation and results, accepted Milonni, P.W., Fugate R.Q. and Telle, J.M.: satisfactorily completed, the detailed for publication in Experimental Astronomy. ‘Analysis of Measured Photon Returns from design activities for the LGSF can be Delplancke F., Ageorges N., Hubin N., O’Sul- Sodium Beacons’, JOSA A, Vol. 15, p. resumed by the second half of next livan C.: “LGS light pollution investigation 217–233, 1998. year, pending approval by the ESO in Calar ALto” Proceedings of the ESO/OSA Murray, J.T., Roberts, W.T., Austin, L. and management. topical meeting on Astronomy with Adaptive Bonaccini, D.: ‘Fiber Raman laser for Optics, Present Results and Future Pro- sodium guide star’, in SPIE Proc. Vol. 3353, grams – Sonthofen, September 1998. p. 330, 1998. References Di Serego Alighieri, S., Bonaccini, D., Oliva, Tyler, G.A.: ‘Rapid Evaluation of d0: the ef- Ageorges N., Delplancke F., Hubin N., Red- E., Piotto, G., Ragazzoni, R. and Richichi, fective diameter of a laser guide star adap- fern M. and Davies R.: “Monitoring of laser A: ‘Adaptive Optics for the Telescopio tive optics system’, JOSA A, 11, 325, 1994. guide star and light pollution”, SPIE Pro- Nazionale Galileo’, Technical Report No. 41, Viard E., Delplancke F., Hubin N. and ceedings 3763, in press, 1999. Astronomical Observatory of Padua-Asia- Ageorges N.: “LGS Na spot elongation Avicola et al.: ‘Sodium Layer guide-star Ex- go, 1995. and Rayleigh scattering effects on Shack- perimental Results’, JOSA A, Vol.11, Fried, D.L. and Belsher, J.F.: ‘Analysis of fun- Hartmann wavefront sensor perform- 825–831, 1994. damental limits to artificial guide star adap- ances”, SPIE Proceedings 3762, in press, Bonaccini, D: ‘Laser Guide Star Adaptive Op- tive optics system performance for 1999. tics Performance Analysis’, ESO Technical astronomical imaging’, JOSA A, Vol. 11, p. Viard E., Delplancke F., Hubin N., Ageorges Report VLT-TRE-ESO-11630-1202, 1996. 277, 1994. N. and Davies R.: “Rayleigh Scattering and Christou, J.C., Bonaccini, D., Ageorges, N. and Le Louarn, M., Foy, R., Hubin, N. and Tallon, laser spot elongation problems at ALFA”, ac- Marchis, F: ‘Myopic Deconvolution of Adap- M.: ‘Laser Guide Star for 3.6 and 8m cepted for publication in Experimental As- tive Optics Images’, The Messenger No. 97, telescopes: performance and astrophysi- tronomy. 14, Sept. 1999. cal implications’, MNRAS 295, 756, Conan, J.M., Fusco, T., Mugnier, L.M., Kersale, 1998. E. and Michau, V: ‘Deconvolution of Adap- Le Louarn, M, Hubin, N., Foy, R. and Tallon, M: ‘Sky coverage and PSF shape with LGS- E-mail: [email protected]

VLT Laser Guide Star Facility Subsystems Design Part I: Fibre Relay Module W. HACKENBERG, D. BONACCINI, G. AVILA, ESO

The advantages of the LGSF fibre re- cation systems to save repeaters. In 3. Scattering at the fibre core-cladding lay approach are twofold: (1) we avoid the fact, much work has been done and lit- interfaces. cumbersome optomechanical relay, and erature is available on the subject 4. Mode conversion from low-loss (2) transfer a diffraction limited beam. The (Tsubokawa et al., 1986; Cotter, 1982). trapped modes to high-loss cladding opto-mechanical relay system would re- However the successful injection of modes. quire a sealed tube for the laser beam op- several watts of continuous wave (CW) 5. Stimulated Raman Scattering tical path to avoid turbulence and dust visible laser power has not been (SRS). contamination, with air-tight sliding joints demonstrated for long fibres. As we are 6. Stimulated Brillouin Scattering at the elevation axis, two servo-controlled concerned with a fibre length of 30 m (SBS). steering mirrors to keep the optical align- for the VLT LGSF system, we have 7. Four-wave mixing between forward ment and several safety interlocks if the been able to achieve this in the labora- and backward Stokes modes. beam has lost alignment. Point (2) is very tory and have also learned a few tricks Points 1 and 6 are the most important important, as much effort needs to be in the process. A fibre relay (Figure 1) causes of losses in our case. spent to ensure a good beam quality. As- provides a very flexible solution and The basic requirements of our fibre re- suming the LGS is an extended object, significantly simplifies the integration. It lay system (Fig. 1) are: the adaptive optics wavefront sensor er- would also be readily applicable to mul- (i) stable throughput above 70 % over ror is proportional to the area of the LGS tiple LGS systems needed for future 30 m for 589-nm CW laser light in the mul- divided by its flux. As it is both costly and high-order AO systems on large tele- ti-Watt regime difficult to increase the laser power, all scopes. (ii) diffraction limited output for the possible efforts have to be taken to min- The potential causes of power losses smallest possible LGS spot size imise the LGS spot size, i.e. the laser in the fibre are: (iii) polarisation preserving. beam quality. Some laser relay systems 1. Coupling losses from the laser beam The latter allows the output of sever- in other LGS projects are even consid- to the fibre core. These include Fresnel al fibres to be combined, and to circu- ering incorporating active or low order reflections and laser to waveguide mode larly polarise the light for the optical adaptive optics in the laser beam relay mismatches.. pumping of the mesospheric sodium optical path, to ensure a good outgoing 2. Extinction losses in the bulk mate- atoms. In order to achieve these re- beam quality. rial. In a 30 m long low-loss, pure fused quirements, three fundamental prob- Incidentally, injection of high laser silica fibre, the bulk losses from scatter- lems have to be solved: power in a single-mode fibre is very ing and absorption are expected to be 1. Efficient coupling of a free-space useful in long distance fibre communi- about 8%. laser beam into a small-frame optical fi-

14 tion is stable, linear and aligned to the birefringence axis. For focusing the free-space laser beam onto the fibre we use an aspheric lens, mounted and pre-aligned to the fibre input inside an athermalised housing. We first carried out an analytical optical design, and then verified numerically the coupling efficiency using CODE V. The experiments reported in the last section of this article confirm the design model. Figure 3 shows the dependence of coupling efficiency on decentre between the optical axis and the fibre axis, for various angles of incidence. For optimum coupling efficiency, the incoming laser beam has to be shaped so that the waist dia- Figure 1: LGSF fibre relay subsystem layout. The fibre input is on the laser optical table, the meter of the focal spot produced by the output is at the launch telescope. The power and polarisation sensors at the fibre end are used external lens exactly matches the fibre to drive the alignment actuators at the fibre input. mode field diameter. In our case this is 15 % larger than the nominal core di- bre. This demands an accuracy and long merical aperture NA = 0.08. For opera- ameter. term stability of the order of a fraction of tion as close as possible at the single- Aligning the coupling unit with respect a micron. mode cut-off wavelength, the maximum to the incoming beam requires three de- 2. Handling of high power densities at fibre core diameter is 5.5 µm for light at grees of freedom. A power and polari- air-glass interfaces. The power density 589 nm. sation sensor analysing the fibre output at the fibre core interface is in the order The linear loss coefficient of fibre with will provide the error signal for a piezo- of several tens of MW/cm2. This is above a core of pure synthetic fused silica is based servo-loop stabilisation of the op- –1 the damage threshold of high power anti- 0.003 m at 589 nm. With antireflection timum injection. reflection coatings (0.5 MW/cm2). A coatings on the fibre ends the theo- In order to prevent the single-mode careful thermal management in the fibre retical maximum throughput of a 30-m fi- contaminating cladding modes (i. e. high- connectors is therefore crucial. bre becomes 90%. Since the damage er order modes that are not guided by the 3. Nonlinear effects. At high power lev- threshold of high-power antireflection core), as well as for an optimal heat re- els the scattering processes become coatings is around 500 kWcm–2 in the vis- moval, the fibre connectors have to be stimulated and influence both effective ible, the air-glass surface has to be en- fabricated out of glass that matches the throughput efficiency and spectral puri- larged at the fibre ends. For this we use fibre core index. A protective and flexible ty of the transmitted light. a fused silica window, either in optical stainless steel sheath is required over the Three simple measures have been tak- contact with the fibre or fusion spliced entire length of the fibre. en to overcome these problems: We have (Fig. 2). To avoid Fabry-Perot effects be- (i) maximised the area of the illuminated tween the fibre ends, the output surface Critical Fibre Input Power air-glass surface, (ii) maximised the ef- has to be wedged. Density fective cross-sectional fibre mode area The state of polarisation at the exit of and (iii) maximised the effective laser line a polarisation-preserving fibre is stable The most important nonlinear loss width. and linear if the input state of polarisa- process in the fibre is stimulated Bril- The limits are set by: (i) the numerical aperture of the fibre, (ii) the require- ment of having single-mode output, (iii) the finite bandwidth of the mesospheric sodium excitation process.

Design of the Single-Mode Fibre and Input Coupling Unit

The input laser beam at the fibre- waveguide is assumed to be in TEM00 mode, i. e. the lowest-order resonator mode. Gaussian beam propagation has to be considered in the design of the coupling optics and their tolerances, so that the mode fields of the input beam and of the optical waveguide optimally match. If we consider a fibre, which is single- mode over its entire length for a given wavelength, the field distribution of the guided fundamental mode can be ap- proximated by a Gaussian. That mode field diameter increases not only with increasing core radius but also with decreasing numerical aperture. The latter is a measure of the light acceptance angle of a fibre. Restricting the waveguide losses due to bending and micro-bend- Figure 2: The glass-air interface area of the fibre is enlarged using a window of the same ma- ing to no more than 1% (minimum bend terial of the fibre core. The window is either spliced to the fibre, or optically contacted. Both radius 10 cm) results in a minimum nu- fibre ends have a fibre window.

15 Energy conservation demands that the amplified backward light (the Stokes wave) is shifted downwards in frequency with respect to the incoming (pump) light, by an amount equal to the acoustic frequency of the material excited. The Stokes line frequency shift is in our case about 30 GHz, and the spectral (or Bril- louin) width of the backscatterd light is ∆νB = 110 MHz FWHM. The effect of SBS is that a fraction of the light injected in the fibre is sent backwards, shifted in frequency and with a characteristic linewidth of ∆νB. SBS is a non-linear effect and its onset is abrupt, once the pump laser power exceeds a certain threshold. This threshold depends on the fibre material and the frequency distribution of the pump. Solving the differential equations for the Stokes and pump intensities in the steady regime, we obtain a critical pump power2 of 720 mW. This is for single-frequency, narrow-band sodium light injected into a 30-m single-mode fused silica fibre whose parameters are Figure 3: Coupling efficiency for the designed optics. The coupling efficiency is plotted against optimised for minimum power density in decentring error between the laser beam and the waveguide axis. Several curves for different the core. In an equivalent polarisation- beam incident angles are shown. maintaining fibre the SBS threshold power is reduced by a factor of two due louin scattering (SBS). This is the scat- photon interaction process. The stand- to polarisation scrambling. tering of light by sound waves that are ing acoustic waves create a regular As long as the coherence length of the created in the fibre by the incident density grating along the fibre. As a pump light is much smaller than the fibre light due to electrostiction1 in a phonon- consequence of the selection rules in length, the frequency-dependent SBS the waveguide, only scattering in the gain coefficient is given by the convolu- 1The type of piezoelectric effect where the induced backward direction occurs in the fi- tion of the pump spectral profile and the material deformation depends on the square of the SBS gain profile. This is a key point. As electrical field amplitude is called electrostiction. bre due to SBS. it means the critical pump power can be increased, if the effective laser linewidth is broadened, for example by means of phase or frequency modulation of the laser beam. If we assume a narrow-band single- frequency CW laser source whose effective linewidth is broadened externally by a simple phase modulation scheme (see below), then the pump spectrum will consist of several longitudinal modes equally separated by a frequency ∆νm corresponding to the modulation bandwidth. The modes of the Stokes wave will be also equidistant, separated by the same distance ∆νm. If again the pump coherence length is much smaller than the characteristic SBS interaction length, i.e. for fibres more than a few metres long, each of the Stokes modes interact with all of the pump modes, and vice versa. Since the interaction strength is determined by the de-tuning from resonance, the interaction bandwidth is limited to the Brillouin linewidth ∆νB. In other words, if ∆νm >> ∆νB then the SBS source polar- isations generated by different pump modes are not phase matched and do not interfere coherently, as each pump mode generates its own Stoke wave. In this situation and with equal power –1 pump modes, the resulting SBS gain is Figure 4: Increase of the SBS threshold pump power, ρeff , under multiline excitation, normalised to the single-mode threshold, as a function of the frequency modulation bandwidth to Brillouin ∆ν ∆ν φ linewidth ratio m / B, and the achievable peak phase shift m. It has been assumed that 2Here defined as that input power for which the the pump coherence length is much smaller than the fibre length, Lf, and that the frequency total SBS power equals 1% of the pump power at modulation is achieved with a single transverse electro-optical modulator. the fibre entrance.

16 reduced by a factor equal to the number of pump modes. This reduction of the SBS gain in case of multimode excitation can be summarised in a factor ρeff, averaged over all modes and normalised to the SBS gain for a single pump mode with the same total intensity and intrinsic linewidth. In general, ρeff will also depend on the relative power distribution within the pump modes. For a frequency modulation scheme with a transverse electro-optic light modulator (see below), the relative strength of pump mode j is given by the square of the Bessel function of order j, evaluated at the peak phase shift φm which can be achieved with the modulator. The inverse of ρeff (or the effective increase in the critical input power) is shown in Figure 4 as a function of the externally applied peak phase shift and the frequency modulation bandwidth to Brillouin linewidth ratio. It can be seen that at φm = 1.5 rad, 2.7 rad, 3.8 rad (etc.) local maxima in –1 ρeff appear. Those extremes correspond to the most equal power distribution within the generated pump modes.

Mesospheric Sodium Excitation Figure 5: Predicted LGS brightness mV in case of laser light relay with an optimised single mode, Efficiency 30-m-long, polarisation-preserving, pure fused silica fibre. The LGS brightness is computed as a function of the frequency modulation bandwidth to Brillouin linewidth ratio, ∆νm / ∆νB, and The effective backscatter cross sec- peak phase shift φm, achieved with a single electro-optic modulator. The parameters for the tion of the mesospheric sodium atoms atmosphere are chosen to represent median values. For the transmission of the launching op- is given by the convolution of the spec- tics 90 % was assumed with the LGS at zenith. tral illumination profile and the sodium absorption profile. We want to max- imise it. 30-m-fibre that is single-mode over its cal input power limits the brightness, in- With a fixed LGS spot size and in- entire length. dependently of the peak phase shift. creasing layer illumination, saturation The throughput of the laser launch- For larger bandwidth ratios, the de- of the atoms can be avoided by broad- ing optics has been assumed to be crease in the overlap integral between ening the effective laser linewidth. 90%. The atmospheric parameters the excitation spectral profile and the Since the width of the Doppler-broad- (seeing, transparency, sodium column Doppler-broadened absorption profile ened sodium D2 absorption profile is density, layer height and temperature) limits the LGS magnitude. The bright- about 2 GHz, there is an optimum ef- assumed for Figure 5 represent median ness gain starts to saturate for optimum fective laser linewidth, which optimises Paranal values, with the telescope peak phase shift values above 2.7 rad. the LGS photon return. In Figure 5 the pointing the LGS to zenith. As can be This, together with a minimisation of the predicted LGS brightness is shown as a seen, the maximum brightness occurs electrical power requirements for the function of the phase modulation pa- at modulation bandwidth to Brillouin laser light modulation, implies an over- rameters, in case of a beam relay with linewidth ratios in the range between 1 all optimum peak phase shift of φm = 3.8 an optimised polarisation preserving and 2. For smaller ratios the lower criti- rad. The corresponding optimum modulation to SBS bandwidth ratio is ∆νm / ∆νB = 1.0. With these parameters the maximum LGS brightness will be limited to 9.5 equivalent V-band magnitude in case of unpolarised excitation.

Experimental Results

Fibre-transmission experiments have been carried out with the ALFA sodium laser on Calar Alto in 1998. The results of the SBS threshold achieved agree well with theory. In Garching, a prototype injection setup has been built with a 532 nm, 5W CW all solid-state laser. The scheme is shown in Figure 6. HPL is the high power laser. For the laboratory experiment this is a diode-pumped Nd:YV04 laser (Coherent Verdi) capable of emit- ting 5.5 W of single-frequency CW pow- Figure 6: Layout of the laboratory set-up for the results reported in this article. See text for ex- er at 532 nm with high beam quality. To planation of the acronyms. suppress any back-reflections into the

17 has been that it is possible to achieve an optimum launching and efficient SBS suppression for the test fibre. More- over, our theoretical calculation and experimental results are in good agreement, which gives us confidence in the reliability of the model of the system. With the current HF amplifier driving the electro-optic light modulator, a maximum peak phase shift of φm = 2.1 rad at ∆νm = 110 MHz can be achieved. At 532 nm this corresponds to a SBS gain reduction factor ρeff = 0.41. In Figure 7 the fibre output and backscattered power are shown as a function of the input power on a 5-m-long, non-polarisation-preserving single-mode fibre. Below 2.4 W input power only Fresnel reflection is observed in the backscattered light. This agrees with the calculated critical SBS pump power of 2.5 W. The slight dis- crepancy between the former values can be explained by the tolerances in the fibre parameters. For input powers higher than 2.5 W and with the laser frequency un-modulated, the SBS back-reflected power increases to 29% of the pump power above the Fresnel reflection. In the spectrum of the back-reflected light the Brillouin-shifted input started to appear at the critical input power. With the modulator switched on, no SBS was observed up to the maximum achievable input power level. This agrees with the calculated decrease of the SBS gain. The Figure 7: Measured fibre output power Po and backscattered power Pb as a function of the input power Pi for a 5-m-long pure fused silica non-polarisation preserving single-mode fibre at modulated throughput at the maximum 532 nm. The cases of unmodulated and frequency modulated input are shown (light modula- input level of 4.2 W is 85 %. For com- tor EOM on/off). For comparison, the extrapolated low power HeNe-laser transmission and the parison the throughput of a low power maximum theoretical throughput is displayed. See text for further details on the setup. HeNe-laser is 88 %, measured with an optical setup optimised for the He-Ne laser, which had a mode quality compa- laser head a Faraday Isolator (FI) is in- beam, the back-reflected light from the rable to the Verdi laser. The difference can stalled near the laser output coupler. M1 fibre entrance and output. These allow the be explained by the larger beam jitter of and M2 are folding mirrors for controlling measurement of the light spectrum (SA), the Verdi laser, which should account for the four degrees of freedom of the laser polarisation (POL), power (photodiodes nearly 3% coupling loss due to concen- beam. PH1-PH4 are pinholes for align- PD 1–4, bolometer B), beam quality tricity errors. The maximum theoretical ment purposes (see below). For the same (BPA) and jitter as well as waist location throughput is about 90% including Fres- reason an attenuator (A) is installed in the and size (beam propagation analyser nel reflection at the uncoated fibre end beam. Its working principle is based on BPA, position sensitive device PSD). The surfaces (8% loss) and extinction (linear Fresnel reflection at counter-moving diagnostics tools are fed with the help of loss coefficient 0.005 m–1 at 532 nm) but glass plates, which ensures minimum windows (W1–3), beamsplitters (BS) and neglecting all other loss sources. The beam position deviation during operation. flipper mirrors (F1–8). Metallic coated maximum fibre output power was stable In combination with the position an- neutral density filters (NDF) are used to during the experiment. A burn-in test with gle of the exit polariser of FI, any at- adjust light levels. All high power lenses longer fibres is scheduled. tenuation between 0 dB (without in- and windows including the phase mod- sertion losses) and 27 dB can be se- ulator crystal are antireflection coated. Conclusions lected. BC and BE form the beam shap- The high power mirrors are dielectric ing optics. These are 2-element 4x coated. To reduce scattering and beam We have demonstrated that a fibre re- beam compressor and expander, re- degradation, all high-power surfaces lay module is feasible. The experimen- spectively. The purpose of these two tel- and the elements in the arm of the beam- tal results agree well with the theoretical escopes is twofold: First, BC reduces the propagation analyser have a surface model. Further work on an optimised, beam size so that the beam can pass quality of better than 60 nm PTV and are end-face-protected, 30-m-long single- through the electro-optical light modula- polished down to 10–5 scratch-dig. The mode fibre is in progress and the first ex- tor (EOM) without clipping. The aperture throughput of the high power beam path periments will be carried out in Garching. of the electro-optical crystal is minimised is 80 %. HeNe and DPSSL are low pow- A burn-in trial on Calar Alto sodium laser in order to relax the electrical power er lasers for alignment purposes. is foreseen in the first half of next year. requirements to drive it. Second, by slight- To allow the optics in the high-power There appear to be at least two pos- ly decollimating BC and BE, the waist lo- arm to thermalise and also for safety rea- sibilities for further increasing the LGS cation and diameter of the beam enter- sons the coarse-alignment of the high brightness within a fibre-fed relay system: ing the focusing lens L1 is controlled for power laser beam is started first with an 1. To combine the polarised output of optimum launching, i. e. mode matching attenuated beam. Figure 7 summarises two fibres at the launch telescope. The with the fibre waveguide. the results on the high power transmis- brightness limit would increase by 0.7 mag. Several diagnostics tools are inte- sion of a 5-m-long single-mode test fibre. 2. To separate the two purposes of the grated for on-line analysis of the launched The conclusion from this first experiment fibre relay, namely flexible transportation

18 of polarised high power light and dif- return flux, a hybrid fibre is under study. power narrow-band laser light in monomode fraction-limited output, into two physical Assuming the same launching and at- fibre’, Electronics Letters, Vol. 18, No. 15, fibre sections: a longer multimode sec- mospheric parameters as for Figure 5, p. 638–640, 1982. tion and a shorter mode cleaning sec- without optical pumping of the sodium Smith, R.G.: ‘Optical Power Handling Capacity of Low Loss Optical Fibres Determined by tion. atoms via circular polarisation, this laser Stimulated Raman and Brillouin Scattering’, The disadvantage of the first solution setup should be able to provide a see- Applied Optics, Vol. 11, p. 2489, 1992. is the increase in complexity, however we ing-limited LGS of brightness V=8 mag- Tsubokawa, M, Seikai, S., Nakashima, T., Shi- know it is feasible. The advantage of a nitude. This would be sufficient for adap- bata, N.: ‘Suppression of Stimulated Bril- two-sectioned fibre is that its design tive optics correction even under poor louin Scattering in a Single Mode fibre would be no longer limited by the critical seeing conditions, pointing at 60 deg by an Acousto-Optic Modulation’, Elec- input power of a single-mode fibre of Zenith angle. tronics Letters, Vo. 22, No. 9, p. 473–475, equivalent length. Instead higher power 1986. levels could be considered. For a sodi- References um laser with 10 W output power, whose Cotter, D.: ‘Suppression of Stimulated Brillouin spectrum is optimised for maximum LGS Scattering during Transmission of High- E-mail: [email protected]

New Pictures from the VLT

Spiral Galaxy Messier 83: This photo shows the central region of a beautiful spiral galaxy, Messier 83, as observed with the FORS1 instrument at VLT ANTU. It is based on a composite of three images, all of which are now available from the ESO Science Data Archive. The three frames were taken in March 1999 through three different filters: B (wavelength 429 nm; Full-Width-Half-Maximum (FWHM) 88 nm; exposure time 10 min; here rendered as blue), R (657 nm; 150 nm; 3 min; green) and I (768 nm; 138 nm; 3 min; red) during a period of 0.8 arcsec average seeing. The field shown measures about 6.8 x 6.8 arcmin and the images were recorded in frames of 2048 x 2048 pixels, each measuring 0.2 arcsec. North is up; East is left.

19 First Images from FORS2 at VLT KUEYEN on Paranal

Top: Three-colour composite of the well-known Crab Nebula as observed on November 10, 1999. It is the remnant of a supernova explosion at a distance of about 6000 light-years, observed for the first time almost 1000 years ago, in the year 1054.

Right: Three-colour composite of the young object Herbig-Haro 34 (HH-34), now in a protostar stage of evolution. It is based on CCD frames obtained with FORS2 in imaging mode, on November 2 and 6, 1999. The object has a remarkable, very complicated appearance that includes two opposite jets (red straight line, down centre) that ram into the surrounding interstellar matter. This structure is produced by a machine-gun-like blast of “bullets” of dense gas ejected from the star at high velocities (approaching 250 km/sec). This seems to indicate that the star experiences episodic “out- bursts” when large chunks of material fall onto it from a surrounding disk. Note also the enigmatic yellow arc to the left, a feature that is still unexplained. Commissioning of the Unit Telescopes of the VLT J. SPYROMILIO, A. WALLANDER AND M. TARENGHI, ESO

In the June 1998 issue of The Mes- newest ideas that had been implement- possible and get every last millisecond senger (No. 92) there was an article deed the day before. Pretty pictures, which out of the loop. Working with a 33 MHz scribing the hectic days and months be- before first light were obtained when the 68040 processor as the central engine fore the first light for UT1. First light was seeing was better than 0.6 arcseconds, was a challenge. However, together with a great public event and of great signif- were banished to be taken at times when Birger Gustafsson, Antonio Longinotti and icance for the project and the organisa- the seeing was better than 0.5 arcsec- Philippe Duhoux (M2 and CCD experts) tion. From a technical point of view how- onds. Then the limit was changed to 0.4 the field stabilisation was brought to an ever, it was a non-event. The telescope arcseconds as seeing was good too of- effective update frequency of 20 Hz. Now had met all specifications for first light ten and test time was being limited. It was was the chance to have some real fun. ahead of time and the weeks before the often said that the seeing was best when Out of the drawer Birger produced the lat- official announcement were more a sit Jason was asleep in the control room. est and best toy. A Power PC processor tight, don’t tell and don’t break it time. Fol- Ivan Muñoz and Anders Wallander set out for our guiding LCU. It was tested on the lowing the big event we were free again to prove this by obtaining a 0.27 arc- control model in Garching, deemed to be to work on the machine. For those with second image in June under those op- good and hey presto! 50 Hz frequency keen eyes looking back at the beautiful erating conditions. A press release on the on the M2. The secondary mirror of the images taken for the first light, flaws could subject – politely avoiding the sleeping VLT is 1.2 metres across. Moving it at 50 easily be found. In fact a number of mes- issue – can be found on the web. Hz is not for the weak of heart. If you got sages and private comments arrived at Antu, or UT1 as it was still known in time on the VLT and you have got some Paranal describing in great detail what we those early days, suffered from wind data, the 'little' mirror at the top of the tel- already knew. A lot of work lay ahead of shake. The telescope was not stiff escope was moving about all the time us. enough to remove the effects of the wind making the focal plane stay put while the The working schedule was set out. on the focal plane. This was not a sur- big thing (telescope) shook in the wind. During the day Peter Gray, Toomas Erm, prise since from the design phase already The weakness of the telescope in wind- Gerd Hudepohl, Marc Sbaihi and Juan it was known that the UTs would be sus- shake rejection was thought to be with us Osorio would fix all the things we broke ceptible to such problems. Before first for life. Toomas Erm would not accept that during the night and continuously im- light Gianlucca Chiozzi and Robert Kar- and continued working on this problem. proved the opto-mechanical and elec- ban had done a splendid job in getting Before first light we only needed the tronic robustness of the system. During field stabilisation to work in a matter of a telescope to point to 10 arcseconds or so. the night Jason would sleep on his sofa couple of weeks. Now was the time to Even for a big machine like UT1 this was while the others tested out the latest and tune it up, remove the lag as much as not very difficult to achieve thanks to the full integration of the pointing model into the tracking software. Now we had the real specifications to meet. Pointing any telescope to 1 arcsecond rms all over the sky is hard enough. Doing it with the UTs of the VLT is an interesting challenge. First we had to stabilise the pointing model which was varying from night to night. It does not take a rocket scientist to work out that if the primary mirror moves inside the cell then the pointing errors of the telescope change. Keeping the mirror in position was provided for with the mirror cell hardware, which permits us to position the mirror in x,y,z as well as the three angles about these axes. By re-positioning the mirror after each pre-set we stabilised the pointing model. A new setup keyword for the telescope was needed and the food theme on Paranal began. When TEL.PSOUP (pronounced pea soup) is set to true the m1 cell Passive SuppOrt moves the mirror UP. Now we had a stable pointing solution and we could start worrying about the actual performance of the telescope. Trying as hard as we might, neither the residual of the fit nor the actual pointing performance could be brought below 1.5 or so arcseconds rms, and often the performance was worse. But then again we asked ourselves why are we bothering with this. The telescope guide probe moves automatically to find a guide star. Why not off- Figure 1: The plot shows the rms tracking error as measured on the telescope encoders for set the telescope position to make the Azimuth on Kueyen. The specification is 0.1 arcsecond and the achieved performance is clear- guide star land where the guide probe ex- ly much better than that. The direction of the arrow indicates the direction the telescope was pected it to be. The pre-setting accura- moving during the measurements. cy would then be equivalent to the as-

21 trometric accuracy of the guide star. Two birds killed with one stone. One was the pointing of the telescope. The other was the reproducibility of the pointing. The latter was to show its true worth later when FORS1 was being commissioned. Now that some basic things were out of the way we could start having fun. Again Gianlucca Chiozzi and Robert Kar- ban were set to work. We want to chop and field stabilise at the same time. Sure they said. It is part of the specs. One beam is always under control while the other is the sky beam that no one cares about. Somewhere in the specifications this had slipped through. Maybe it came from the days of single-element IR detectors. In any case, the new more complex specification was given out. As- suming the guide star is going to land on chip in both beams of the chop then we wish to field stabilise on both beams. Back to the drawing board. Philippe Duhoux to the rescue with a port of the DIMM software to the guider and again our wish was fulfilled. ISAAC folks would be happy. But what about the comet and moving-target community? Of course the VLT can track on a moving target. But could it guide on it as well? So the challenge was set. How would this be done? Back to the original design to see what was thought to be the way. Move the guiding box on the CCD with the appropriate speed and everything will be OK was the theory. But why do that? Why not tell the Figure 2: The top left plot shows the measured centroid for a star during successive 30 sec- guide probe to simply track its star. Then ond exposures taken during a 3.5 hour track. The difference in the centroid between the first if the centre moved or not was not a prob- image and all successive images is the measured error in tracking. The three other plots show lem since the guide probe would happi- the error in tracking as a function of altitude, adapter angle and hour angle. The telescope ly hunt down its guide star. It actually crossed the meridian during this test and the peak error was below 0.15 arcsecond during turned out that Birger Gustafsson had al- the entire experiment. ready coded almost all of this simply to be able to pre-set the probe to the guide about this wind shake thing. Gerd Hude- lutions by Stefan and Rodrigo and the star. All he had to do was to keep the pohl and Pierre Sangaset would fret ASM was up and running. Now the ra- probe alive at all times. Easier said than about the cell while Gustavo Rahmer was dio need not sound all night long with the done but, as usual, nothing is more fun finally about to get some serious work to call: “Julio! What is the seeing?” than the impossible. So having this work- do. Thanh Phan Duc would come and go Before putting FORS1 on to the tele- ing, we could now chop and field stabilise and let us all know exactly how many cell scope we had to satisfy another client. while the telescope tracked a moving tar- operations we had made and would fix Bruno Leibundgut together with Mark Fer- get. Unfortunately, in doing this we whatever needed fixing. It was like having rari and Eline Tolstoy arrived for science broke the combined offset mechanism. a policeman checking up all the time. Ri- verification with the test camera. Brilliant Traditionally the probe is offset in the oth- cardo Schmutzer moved between Garch- plans were laid out, astronomical chal- er direction to the telescope. By making ing and UVES and Paranal and the test lenges and problems to be solved. The the probe live when the telescope offset camera without missing a beat on either first couple of nights went fine. In fact, a the probe would automatically follow its of them. FORS1 arrived on the mountain new image quality record was set at 0.26 own star. After a bit of soul searching we and immediately aroused suspicions. Did arcseconds, the image taken with the realised that this was actually the better they really think that big yellow thing ADC mounted. Then the weather way of doing things. The autoguider was would fit under our delicate telescope? changed, the seeing became poor, the promoted above the tracking software as Meanwhile the final phases of the com- control room was not a happy place and the master and we were back in business. missioning of another telescope on before you could say “bad luck” the two While all of this was going on, from Paranal were taking place. The ASM (As- weeks of SV were over. Bruno took a nowhere came a couple of Nasmyth tronomical Site Monitor) was getting to the copy of the data with him and the video adapters and were miraculously mount- phase where it could be left alone. Ste- conference to Garching fell silent. ed on the telescope without interfering fan Sandrock and Rodrigo Amestica Soon afterwards The FORS1 team led with our work (we suspect a phantom would try to make our robotic telescope by Professor Appenzeller together with Francis Franza and his twin Paul Gior- work. Anders expressed wishes that the Gero Rupprecht as the ESO responsible dano had something to do with this but UTs be so simple to operate. Switch it on put their pride and joy on the Cassegrain Martin Cullum denies all knowledge as and the ASM will know when to open and focus of UT1 and then had to take it off to how this miracle occurred). The engi- when to close, where to find a star and again due to a broken cooling pipe inside neering crew was growing by the day. what to measure. Not too hard a re- the adapter which dripped water inside George Harding, the two Pablos (Gutiér- quirement really. Only the ASM was hav- their instrument. We apologised, helped rez and Barriga) and the formidable Patri- ing problems finding alpha Cen let alone to clean up the mess and on went FORS1 cio Ibanez all started putting their stamp any other star. A bit of investigating on the again. The moment of excitement arrived. on things. Toomas Erm kept worrying pointing problem and a few ingenious so- The shutter was opened and the first im-

22 scope. Taking advantage of the absence of the head of commissioning, Michael Loose, our architect on Paranal, trans- formed the control room into Beirut at the height of the Lebanon troubles. The floor and the ceiling disappeared. If a small thermonuclear weapon had gone off it would have made a smaller mess. In spite of this, Jean-Gabriel, Jens Knudstrup and a few other brave souls continued working, wearing surgical masks to limit the amount of dust entering their lungs and the ear-plugs to reduce the noise level to that of a firing range. All was brought together just in time for the end of the commissioning of ISAAC and we were ready for the operations dry runs which were scheduled for March. In the mean time two big events were on the horizon. UT2 (soon to be renamed to Kueyen) was getting ready for first light and the inauguration of Paranal was also rapidly approaching. Without commissioning really noticing, Peter Gray and his crew, with some help from the usual sus- pects in Garching – Bruno Gilli, Max Kraus and Fabio Biancat Marchet come immediately to mind – had another 8-metre telescope ready to go. Cells moving up and down the mountain, mirrors coated, servo loops tuned and all the things Figure 3: Measured pointing errors on 140 astrometric stars covering the whole sky. The meas- necessary to make a telescope ready ured rms pointing on this night was 0.85 arcsecond. somehow were all done without slowing UT1 work. Two telescopes running be- age obtained. The image quality was them we put water inside their instrument fore the inauguration and a nice image quickly derived to be 0.6 arcseconds. A again, but by now we were all experts at to show from Antu would make a good quiet note of satisfaction was felt across dismounting and cleaning FORS1 up. showing. Massimo in a typically chal- the control room and now the real work Our apologies were gracefully accepted lenging order stated that nothing worse commissioning FORS1 could start in again and we started to work together to- than a third of an arcsecond for the im- earnest. A separate article on the FORS1 wards the final integration of the instru- age from Antu would satisfy him. On the commissioning has already been pub- ment into its new home. Unfortunately the night before the inauguration, UT1 had lished, so not much more will be said here. weather did not co-operate and most of already been handed to Science Oper- In the mean time ISAAC was sitting in the second commissioning run was lost. ations for their dry runs. The telescope the integration lab waiting to be installed We scheduled another run for January was theirs. What could they do? Just to on the telescope. Jean-Luis Lizon had and with the help of Thomas Szeifert, set the standard for the future, Hermann been working furiously under difficult con- Bernard Muschielok and Wolfgang Boehnhardt and Roberto Gilmozzi were ditions to complete the re-integration of Gaessler the instrument was Paranalised. at the controls with Massimo directing that the instrument. In the couple of weeks be- Following this it was handed to Fernan- FORS1 be switched to high-resolution tween FORS1 and ISAAC we checked do Comerón to system test. Of course, mode since the seeing was good. It is a out the Nasmyth A focus and verified all Fernando kept on finding things we had well-known fact that the presence of the was well. ISAAC went on to the telescope missed. Roberto Gilmozzi would agree director on Paranal improves the seeing. on schedule and also produced beauti- with him and then we had to change The combination of all of the effects de- ful images straight away. Finally we had everything again. It was fun. scribed resulted in a 600-second I band an instrument that could exercise the tel- Time had come to welcome ISAAC image with quality better than 0.25 arc- escope in all the weird and wonderful back. Remember how well the telescope seconds. Now that the first goal was met ways that only infrared astronomers can and instrument had behaved in the first we could have some extra fun. Lothar find. Alan Moorwood, Jean-Gabriel Cuby commissioning run. Well, this time things Noethe and Stephane Guisard had been and Chris Lidman all worked furiously at were not to be so easy going. The UTs’ labouring for some time to make the tel- characterising the instrument. Our first fel- direct drive motors (big fancy magnets escope spell its own name. Anyone they low on Paranal, Monica Petr, arrived with a protective metal cover) move the said could do it moving the telescope around this time and immediately start- machine silently and very elegantly around until the star trails spelt out VLT. ed learning and driving both instruments around the sky. On side A (not the ISAAC Even the NTT managed that. So they and telescope. Joar Brynnel finally could side) one of the covers came off and got took charge of the shape of the primary see all his electronics work used for as- crunched inside the motor. Patricio mirror and they made stars whose tronomy rather than just tests. Manfred Ibanez, Pablo Gutiérrez, Gustavo Rah- shapes spelt out VLT! True control of the Meyer and Gert Finger got the last ounce mer and Hans Gemperlein worked non- shape of the primary was demonstrated of efficiency out their IRACE system to stop for a week to disassemble one whole beyond a doubt. read the ISAAC detectors. Peter Bier- side of the telescope and put it back to- On Kueyen as soon as we had a pri- eichel was always at hand to make the gether again. The badge of honour at the mary and a secondary mirror installed we instrument software that little bit better. time was a piece of yellow duct tape, had a look. It was not going to be as easy Nicolas Devillard demonstrated the lat- which was worn with great pride for the as Antu but it would work. Lothar Noe- est pipeline capabilities. duration of the work. You might have the, Stephane Guisard and Roberto It was Christmas time by now and the thought that the control room would have Abuter set about their merry way to make FORS1 crew were back. To welcome been a quiet place with a broken tele- the thing make stars that looked round.

23 Krister Wirenstrand sorted out the point- to the fore. In any normal telescope ob- sive pace, the stars were not coming out ing with a bit of help from a big piece of servatory Antu would have been released round but rather long like sausages. Here paper being held at the Nasmyth Focus and the improvements coming from was a problem that had to have a solu- and using a real bright star. Two tele- Kueyen would have been implemented tion. So we all worked on it. Krister scopes working and the best ever image. in some distant future. At Paranal the cy- changed the tracking software. Stefan We now felt that the observatory could cle still has some way to run. changed it back and then back again and be inaugurated. On Antu, for each activity undertaken once more. Rodrigo and Anders meas- Before the telescope was released into by commissioning, a test procedure was ured in ever more imaginative ways. In full operations we threw away the old soft- supposed to be followed documenting ex- parallel, continuous improvements in the ware. Giorgio Filippi came down with a actly how each test was to be performed. system were made, all contributing in their present from Garching. A Y2K compliant This was to be a self-documenting sys- own ways in attacking this problem and system had been fully tested on the tem with the procedure also providing the other annoyances. Mario Kiekebusch au- Garching control model and was installed final documentation for the test. In some tomated the adapter calibration proce- on all systems on Paranal. Carlos Guirao cases, such as tracking and field-stabil- dures and gave us smarter secondary brought some new data flow computers isation maps, this was done while for oth- guiding. Ivan Muñoz made the pointing and Michèle Peron, Nick Kornweibel, ers the procedures were developed dur- modelling trivial by automating it and in- Michele Zamparelli, Gerhard Mekiffer, ing the commissioning, as we better un- troduced us to Pinky and the Brain. Fran- Klaus Banse, Miguel Albrecht, Paola Am- derstood what needed to be done. On cisco Delgado kept improving the active ico, Reinhard Hanuschik, Palle Møller, Kueyen this documentation effort was optics software to allow us to better con- Michael Rauch, Almudena Prieto, Dave much improved and expanded. Also au- trol the mirror position. Sausage meas- Silva and Maurizio Chavan all spent quite tomation of a number of tests was un- urements were made, analysed, models some sleepless nights on Paranal en- dertaken. The use of the engineering data and simulators written. The sausage suring the data-flow system was fully op- stream was expanded dramatically to pro- problem was reduced by a factor of 10 erational before and immediately after vide input for preventive maintenance but and is now an operational annoyance. Antu was released. also to better understand the system. The We keep working on it. Same with the Somewhere earlier in this story VLT control-system engineering-data pointing that has been measured as good Toomas Erm was still worrying about this stream includes for example every aber- as 0.85 arcseconds all over the sky. wind-shake problem. So, after a long ration measured by the system. In any Roberto Rojas has inherited the thermal think about it all, he asked if Stefan San- given month, as many as 20 thousand ac- control software, from Roberto Abuter drock could give him a helping hand with tive-optics corrections are made. Look- who escaped to Enteferometry [sic], and an idea he had. A couple of days later a ing at the logs and using parallel meas- put it into full operation. This promises in new version of our tachometer was im- urements with the test camera wavefront the medium to long term to provide a sig- plemented in software and the enchila- sensor, we identified some cross-talk be- nificant improvement in how we manage da was born (a soft taco being an en- tween defocus and spherical aberration. the telescope environment. chilada). Wind shake is not a thing of the The effect was small resulting in addi- This article is being written as UVES past but Kueyen is now stiffer than we tional noise in the system of order 0.03 ended its first commissioning run on had ever dreamed of. Enchilada gave arcseconds in the average image quali- Kueyen. UVES is looking good. Very way to Enchilada turbo and by now the ty delivered. Lothar was looking at the good! FORS2 is already mounted on the azimuth axis of Kueyen was exceeding data in Garching and promised a solution telescope and has seen stars. Kueyen is tracking specifications by over an order by the birth of his son. On schedule on well on its way to be released on sched- of magnitude and altitude was in great the 3rd of August the problem was iden- ule for Period 65. Norma Hurtado and shape also exceeding specifications by tified and corrected for. A baby boy was Julio Navarrete are driving the telescope a factor of 2 under most operating con- also born on the same day. The cross- and providing valuable operator feedback ditions. The windscreens of the enclosure talk was identified as being due to an in- to the commissioning team. Baring un- have been improved and tested and correct sign in the force setting on the pri- foreseen events in 2000 both Melipal and retrofitted thanks to the work of Martina mary. The improvement in the stability of Yepun will have their first lights and we García (who somehow also managed to the defocus was obvious (the rms value have new ideas for further improvements have a baby while all of this was hap- of the aberration was reduced by 500 to the system. pening) and Erich Bugueno. Juan Oso- nanometres). The value of the engi- To know who has made all of this pos- rio fixed things on the electronics of the neering data stream cannot be over- sible, access the web and get a list of all enclosure and German Ehrenfeld never stated. ESO staff irrespective of location, posi- seemed to stop making things and mov- Continuing on the food theme (pea tion or function. We would like to thank ing big pieces around. Nelson Montano soup, enchilada) we started worrying Krister Wirenstrand personally for his tire- is now our hydrostatic pad man and rou- about the sausages the telescope was less efforts towards building and com- tine preventative maintenance started in producing. Close to the zenith and the missioning the telescopes. earnest. The benefits of having more than meridian, when 430 tons of glass and one telescope to work on were coming metal are spinning around at an impres- E-mail: [email protected]

The Science Verification Plan for FORS2 and UVES at UT2/Kueyen A. RENZINI and P. ROSATI, ESO

1. Introduction of Antu (UT1) and its instruments • FORS2 SV: February 28 – March (www.eso.org/science/utlsv/). SV ob- 10 (New Moon is March 6). Raw, cali- The Science Verification (SV) obser- servations are now planned as fol- bration and calibrated data will be made vations for FORS2 and UVES at Kueyen lows: publicly available to the ESO communi- (UT2) will follow the same approach and • UVES SV: February 6–17 (New ty as soon as the reductions are com- policy as already described for the SV Moon is February 5) pleted and the data prepared for re- 24 lease. The SV Teams plan to release found, including target lists, modes and adding deep U and Gunn z images of the UVES data by March 31, 2000, and the planned exposure times. cluster. FORS2 data by April 30, 2000. • Hα Rotation Curves of Dwarf Observations will be made in Service 2. UVES SV Plan Galaxies. This is a “bad seeing” pro- Mode with SV Astronomers on Paranal Three main SV Programmes have gramme. Long slit spectra should allow for part of the time. SV observations been planned: to constrain the mass distribution at large may actually be taken from February • A Spectroscopic Survey of Inter- radii in a set of three dwarf, dark matter 6 through March 31 in a flexible sched- mediate Redshift QSOs. Five out of a dominated galaxies. uling, depending on the needs of oth- list of 8 QSOs with 2 <~ z <~ 3 will be ob- • Kinematics of Previously Mi- er commissioning and training activities served. These observations are meant to crolensed Stars in LMC. Over two that may arise. be the first step towards a public survey dozens microlensing events have been SV observations will be conducted in of QSO absorbers which aims at creat- detected so far towards LMC. The plan coordination with the PIs of the two in- ing a large and homogeneous dataset to foresees to take spectra of at least 10 of struments, the Instrument Scientists, and allow the study of the intergalactic medi- the stars that were microlensed, in order the Kueyen (UT2) and instrument Com- um in the redshift interval z = 1.6–3. A to obtain their radial velocity with an ac- missioning Teams. team co-ordinated by ESO will submit a curacy of ~ 10 km/s. Constraints on the The scientific programmes have been proposal to carry out such a Public Sur- nature and location of the lensing objects selected in such a way as to produce first- vey in Period 66. could then be derived. quality data for cutting-edge science in • Abundance Ratios of Extremely • Lyα emission at high redshift. The areas of interest for a wide section of the Metal Poor Stars from the Hamburg NTT Deep Field will be imaged with a cus- potential users in the ESO community. ESO Survey. At least 5 stars with [Fe/H] tom narrow-band filter (FW = 7Å) search- Moreover, programmes that require very <~ –3 will be observed. The high S/N spec- ing for Lyα emission from large-scale good seeing are complemented by oth- tra will allow to determine the abundances structure centred on a damped Lyα sys- ers that would still be well feasible in less of a number of interesting elements, such tem at z = 3.2. demanding conditions. as Li, α-elements, and s- and r-process • Near-IR Imaging of the NTT DF. The SV Teams includes the following elements. Imaging in IF915 and Gunn z filters will en- scientists: Joao Alves, Stephane Arnauts. • Chemical Abundances in LMC sure the completion of the deep multi- Jacqueline Bergeron, Tom Broadhurst, Clusters in a Wide Age Range. A set colour database of this deep field, thus Stefano Cristiani, Chris Gorski, Vanessa of Red Giants (16.5 < V < 17) will be allowing a photometric redshift search for Hill, Richard Hook, Rodrigo Ibata, Tae- observed in seven LMC Globular Clus- very distant galaxies. Sun Kim, Markus Kissler-Patig, Benoit ters with ages from 0.1 to ~ 14 Gyr. This • Spectroscopy of High Redshift Pirenne, Mario Nonino, Francesca Pri- should allow a detailed reconstruction of Galaxies. Distant, photometric redshift mas, Michael Rauch, Alvio Renzini, the chemical enrichment history of LMC. selected galaxies to IAB ȃ 25 from pub- Piero Rosati, and Eline Tolstoy. lic NTT data will be observed in MOS The following list of SV programmes 3. FORS-2 SV Plan mode, possibly using the mask option. has to be taken as indicative. In par- The following programmes have been • Abundance Measurements in Gi- ticular, some additional relatively small planned for the SV observations of ant Ellipticals. MOS and long-slit spec- programme could be accommodated, de- FORS2: troscopy for the abundance determina- pending on the actual observing condi- • Deep Imaging of the Cluster of tion of globular clusters and the dwarf tions on Paranal. Any modification will be Galaxies MS1008-1224. This z = 0.3 galaxies surrounding NGC4472 and of promptly reported in the SV WEB pages cluster was extensively observed for the galaxy itself out to ~ 3Re. (http://http.hq.eso.org/science/ut2sv/), SV of FORS1. With FORS2 SV one in- where more detailed information can be tends to complete the database by E-mail: [email protected]

The First Six Months of VLT Science Operations R. GILMOZZI, ESO

Science Operations started at UT1 ISAAC had just had to be taken out of It is a testimony of the excellent work Antu on April 1, 1999. This article is a the instrument to be repaired, so ISAAC of all the people involved that on April 3, brief account of the first six months could only work in imaging mode; the the wind and humidity having abated to (Period 63). As the statistics below will decontamination of the FORS1 CCD the more normal Paranal values, the show, operations have been fairly suc- had not resolved the infamous “quad- first OBs could be executed on the sky. cessful. Visitor and Service modes rangle” problem, and the CCD had to Paranal was finally a full-fledged astro- were intermixed during P63, in a 50/50 be taken out to be baked more thor- nomical Observatory! ratio for the available time (i.e. dis- oughly, so FORS could not be used. From then on, the month spent in counting Guaranteed time, which is by And the weather was the worst seen in ‘dry runs’ to acquaint ourselves with definition in Visitor mode): we had 62 the last year (at least by the Science the telescope and the instruments nights in Visitor, and 60 in Service. To Operations staff): foggy (!), cold and showed its results: it was immediately this one should add 10 Service nights very windy. April 1 and 2 were spent clear that the efficiency of the opera- for the Calibration Plans, and any time with the Paranal Engineering depart- tions was quite high, even if some pro- allocated but not used for technical ment, the Garching Optical Detector grammes were not as efficient as they time, target of opportunity time and di- Group and other Instrumentation Divi- could have been (the fact that ESO had rector discretionary time (so that in the sion staff working overtime on one side decided to absorb the overheads being end, more than 85 nights were devoted to bring ISAAC up in imaging mode so possibly a reason for this), the seeing to Service Mode observations and cali- that the first Service Mode episode of was co-operating very nicely (though brations). the Period could start, on the other to not all through the Period, alas), the in- Period 63 started under less than decontaminate the CCD in time for the struments were behaving very well. promising omens: the mask wheel of first dark moon and the first FORS1 run. Total downtime, both the weather and

25 the technical flavours, was less than 20%. Ten days later the next challenge: the first Visiting Astronomers started coming to the mountain. I should probably explain that Service Mode and Visitor Mode, as seen from the Operations point of view, differ only in who makes the scientific decisions about ‘what to do next’. A Paranal astronomer is always present to run the instrument, to advise the Visitor and to help preparing the OBs beforehand. Therefore the Visiting Astronomers can devote their time to analysing the data and decide on the observing strategy. The experience with Visitor Mode has been very positive, the interaction with the Operations Astronomers on the whole resulting in satisfactory observa- Figure 1. tions (or at least this is what we gather from the end of run reports!). We have learned a lot during the first ed by the oversubscription, but in some active optics correction). For those of six months, and we have also made a cases it has been very difficult to decide you who have not yet had a chance to few mistakes. what to observe. This problem is more see the active optics in action, it is One thing that we have to clarify bet- acute at the end of a period (again, we something really impressive: once you ter to our “proposers” in the future is are thinking of ways to improve this). acquire a new field, the guide probe that a successful Service Observing Particularly demanding was the August goes automatically to a guide star, ac- Mode (SM) needs to include pro- full moon episode when ISAAC was not quires it, starts the image analysis and, grammes designed for various types of available for 7 nights, and we had to usually about 30 seconds later, sends conditions: if everybody wants 0.4 arc- use FORS1 with very few programmes the first correction to the mirror actua- second seeing, photometric sky and be that made sense to execute. tors: at this point you see the guide star one day from new moon, it’s very diffi- Sometimes there is no alternative but shrink to the outside seeing value, and cult that we can satisfy all requests... to violate some constraint: when we you can start your science exposure In Period 63 there were two main is- had to do that, we always tried to com- (see http://www.eso.org/outreach/press- sues that caused some confusion with pensate in some way (for example: rel/pr-1999/phot-34-99.html). our users: the oversubscription (espe- higher than requested moon back- This is the most visible difference cially of category A) SM programmes, ground usually meant that we would ob- with the way we used to do astronomy: and the constraints set. serve in better than requested seeing, from this moment on, the telescope is The first is not a problem per se: to improve the contrast, or in the kept at the best focus, with the best op- oversubscribing the available time adds longest possible wavelength, where the tical figure for the pointing, all in paral- to the scheduling flexibility and makes effect is smallest, or both). Not always lel with the science observations and for an efficient operation. Of course, it this has been possible; ranking has without requiring any additional time. also creates unhappy users, who have played a role (if a choice had to be The typical overhead to move to a new to submit their OBs without guarantee made, we tried to avoid affecting very field, acquire a guide star, and start the of execution. The OPC ranking, which highly ranked programmes). But we active optics is about 5 minutes. is our primary decision input, helped in have found ourselves in situations (for- selecting among programmes compet- tunately few) where it was either ob- Downtime Statistics ing for the same conditions. serve something outside the con- In P63 we were not helped by the straints or keep the telescope idle. One way to “measure” how well we time being assigned in terms of shutter Discussions are ongoing within ESO, did in Period 63 is to compute the total open time: of the 600 available hours in and with the community, to minimise downtime. The technical downtime was Service Mode, 720 were allocated in these problems in the future. Users can 13,561 minutes (11.4% of the total P63. With the measured values of over- help too, by not requesting indiscrimi- available time). This statistics includes head (~30%) and downtime (~20%), we nately conditions that happen only a any observing time affected by the were oversubscribed by (1.3×720)/ small fraction of the time. technical problem (e.g. a problem (0.8×600) ~ 2. Add to this that weather In the following I would like to offer which can be fixed in 5 minutes but variability sometime caused the condi- some statistics for Period 63. Some of which has caused the loss of a, say, 30 tions to change during an exposure, so this statistics comes from the engineer- min exposure would count as 35 min that it had to be repeated, adding to the ing data stream that is recorded every downtime). oversubscription. We are actively study- night, much in the same way that satel- The weather downtime was 12,700 ing ways to decrease the impact of this lite engineering data are recorded (in- minutes (10.7% of time). This includes problem; in P63 we were lucky enough deed, this has proven to be among the clouds (5.6% of time), high winds that part of the technical time (and best tools produced by the Commis- (>18m/s) and high humidity (or combi- some of the discretionary time) were sioning team to understand the per- nations). Considering that P63 was dur- not used, so that we put it into the SM formance of the telescope and instru- ing the southern winter, this percent pool, and were able to carry out more ments). loss does not seem surprising. observations (see the statistics below). In total, therefore, we lost 22.1% of The other problem we had is the con- Performance of Antu the time. straints set: the number of programmes that could be observed when the During period 63 there have been Weather seeing was > 1″ or the moon was up 3907 presets (21 per night), 22,491 off- (for FORS), or for nights with cirrus was sets (6 per preset) and 121,537 image Of the 163 nights that were not lost very small. In part this was compensat- analyses (98% of which resulted in an to bad weather, 118 (72%) were ap-

26 parently photometric, 18 (11%) possibly cially in visitor mode, it is an unlikely oc- (6/0/0) are category A open pro- photometric, 17 (11%) had thin cirrus and currence. grammes that the Director General has 10 (6%) had thick cirrus. We are currently allowed to carry over into Period 64 so finalising the analysis of the standard Operations Statistics that they can be finished. stars observations to confirm the number The success of the first six months of of “true” photometric nights. ISAAC was used 73 nights, or 40% operations is the result of the commitment The statistics on the seeing is based of the time. FORS1 was used 110 and professionality of a very large num- on 116,761 ASM data points, each av- nights. This reflects in part the instru- ber of people, both in Europe and in Chile. eraging one minute of measurements. mental problems described above, in Naming all of them would look very much The median seeing in P63 was 0.76”. The part the actual time allocation (or the like the ESO internal organigram! How- seeing was better than 1” 72% of the time, user preferences). ever, I would like to mention at least the and better than 0.5” 13% of the time. Once one discounts the downtime, Science Operations Team at Paranal You may have heard that sometimes there were 1484 available “observable” (now under the leadership of Gautier the seeing measured by the instru- hours. Of these, 651 were used for Mathys), the Paranal Engineering De- ments on Antu is better than the outside FORS1 and 403 for ISAAC. This results partment (Peter Gray), the User Support seeing. Discounting wavelengths effect in a total “shutter open” efficiency of Group (Dave Silva) and the Data Flow (i.e. yes, it is true that ISAAC with its 71% (see Figure 1). In period 63, the ar- Operations Group (Bruno Leibundgut) for 0.15″ pixels sometimes undersamples chive “ingested” 19,794 ISAAC frames their dedication and enthusiasm, and for the atmospheric PSF), we have indeed and 24,039 FORS1 frames. their crucial contributions. Jason Spy- seen this effect. It is however fairly rare, In Service Mode, we had 75 accept- romilio and his team delivered fully com- and needs special circumstances (in ed programmes (40 in category A, 19 in missioned telescope and instruments. particular, the appropriate wind speed, B and 16 in C, or 40/19/16 for simplici- Massimo Tarenghi as director of Paranal a pointing such that the wind can flush ty). The completion statistics is 34/16/ provided oversight and direction to the the primary at the right angle, and an 11 (of which 13/10/8 are partial comple- whole process. excellent seeing to start with). On some tions, usually meaning that most of the Special thanks to Jason, Gautier, occasions, in service mode, we have programme was carried out, the re- Dave, Chris Lidman and Jose Parra for had enough programmes in the queue mainder being impossible to observe providing the statistics in this article. that we could choose to observe in the because of target RA or non realised right direction, and so took advantage constraints etc). The programmes not of this. In general, however, and espe- initiated were 0/3/5. The remainder E-mail: [email protected]

Configuration Management of the Very Large Telescope Control Software F. CARBOGNANI, G. FILIPPI, P. SIVERA, ESO, Garching

1. Introduction client server mechanism. In our proj- modification must be started, imple- ects, there is typically only one person mented and tested, then archived. One of the elements of the success at a time dealing with a specific part. A During this period the module is locked. of the development, integration and commissioning of NTT, VLT and the at- tached Instruments has been the Con- figuration Management of the related Control Software. It has been based on the Code Archive and the VLT Software Problem Report (VLTSPR) procedure.

2. Code Archive

The code archive supports code configuration during development and integration on geographically distributed sites. The design keys are the following: • The configuration item is the software module. This allows a reasonable but still simple flexibility in system configuration (a system is made up of 15 to 100 items, identified by their name and version, corresponding to 2000 to 20,000 files). Each module is a set of files organised in a fixed directory structure and has a unique name. Figure 1 gives the modules divided by VLT common software and applications. • There is only one central archive. Users get local copies using a simple Figure 1.

27 Figure 2.

• To allow experiments or patches on older versions, branches are also supported. The code archive is implemented using RCS and a set of ad hoc programmes and scripts (cmm) implementing the client-server interaction and the user interface. The archive is physically located at the headquarters and used also by teams in Chile and several institutes. Figure 2 gives the accesses per month for modification and for read- only mode. Currently approximately 60% of the accesses are from outside the headquarters and 25% from non- ESO sites. It can be remarked that the system was able to deal with more than 500 archive and 2500 read-only month- ly accesses. Figure 3 shows in detail the access made by non-ESO sites. The central archive is not only the software repository, but it is also a management tool. In archiving a module, the developer tells all the other people “Hey, there is a new consistent and tested set Figure 3. of files that you can use!”. This is very useful when several teams operating on archive, the integration responsible can version is normally enough to decide different time zones do development and see that new items have been produced. whether to take the new version or not. integration. Comparing the current con- A glance to the comments stored by the In this way NTT, VLT and instruments figuration against the status of the developer to qualify the newly archived have been developed, integrated and

28 • Responsible works on problem • Responsible can close the SPR (must add a final remark on it) • People can add comments any time. At present there are about 120 names in the user database, 75% ESO users (both Europe and Chile), the remaining 25% from 13 external projects, some project have more sites. Figure 4 shows the present distribution of the VLTSPR users. In Figure 5 the trend of the SPR archive over the years is shown. The number of open SPRs has always been kept under a physiological limit (about 500) that corresponds to what we are able to treat between two releases. Figure 6 reports the distribution of the SPRs per area in 1997 and in 1998. As the project evolved, SPRs concerning the common software are decreasing, while the ones for the application part are going up, which is a sign of the integration Figure 4. activity that is now taking place. 4. Conclusion

Tools and Methodologies developed within the VLT Software Engineering Group have proved to be effective in supporting the VLT Software development. They can be seen as a concrete guideline and reference for all ESO internal projects and are proposed as baseline for all external collaborations.

This article is based on reports presented at ICALEPCS 99: [1] G. Filippi, “Software Engineering for ESO’s VLT project”, ICALEPCS 93. Figure 5. [2] G. Filippi, F. Carbognani, “Software prac- tices used in the ESO Very Large Tele- scope Control Software”, ICALEPCS 99. commissioned on top of mountains at a Responsible Person is appointed for the 2500 m in the middle of a desert, but in- problem E-mail: [email protected] volving people in several countries on both sides of the ocean. The Central Archive is hosted on a RAID system level 5, with hot-standby disk, and backed up in two different ways every day. This year the downtime of the server has been less than 8 days, corresponding to 2%, mainly during weekends. We are developing a better client-server mechanism and we plan to make the cmm software available on the public domain.

3. VLT Software Problem Report (VLTSPR)

The VLTSPR System is meant to be used by both internal and external users of VLT software to report errors in code or documentation or to propose a change. VLTSPR is built using the commercial tool Action Remedy (c) and has a Web Browser interface. This is the basic workflow of the system: • Problem submitted, depending on the subject, some people are notified immediately • The SPR is discussed in the Software Configuration Control Board meeting and Figure 6.

29 Java for Astronomy: Software Development at ESO/ST-ECF M. DOLENSKY1, M. ALBRECHT 2, R. ALBRECHT 1, P. BALLESTER2, C. BOAROTTO2, A. BRIGHTON 2, T. CANAVAN 2, M. CHAVAN 2, G. CHIOZZI3, A. DISARÒ2, B. KEMP 2 1ESA/ST-ECF; 2ESO/DMD; 3ESO/VLT

Overview which a Java Virtual Machine (JVM) ex- more concepts and advantages as well ists (e.g. Windows, Solaris, Mac OS, and as the weak points of Java technology. The first part of this article discusses Linux). Some web browsers come with the Java computing language and a num- a built-in JVM. More advantages and supported ber of concepts based on it. After learn- • JavaBeans Component Model = built- concepts ing why Java is best suited for many ap- in support for code re-use: Java incor- • Remote Method Invocation (RMI): plications in astronomy including ESO’s porates JavaBeans which is a software RMI allows the execution of tasks on oth- JSky initiative and certain Web services component model. The component mod- er machines. there is a second part containing a col- el describes how to build an application • Jini: Jini defines a protocol for com- lection of application briefs from software from software components. In contrast to munication with devices. A device is, for projects at ESO and ST-ECF. Appended software libraries Beans provide an ad- instance, a disk drive, a TV, VCR or cell- is a list of contact points in house (Table ditional interface in order to support their phone. With Jini there is no need for sep- 1) and an applet gallery (Table 2). re-use and customisation. The Java- arate driver software. It's a very general Beans concept is widely supported by de- implementation of plug and play capa- 1. Why Java? velopment tools. bilities. • Common look and feel: There is built- • Enterprise JavaBeans (EJB): EJB is Here are the key features [1] that the in support of powerful GUI widgets a specification for server side compo- authors felt were the most important ar- (Swing classes) and common look and nents based on JavaBeans. EJB com- guments why Java was the first choice feel on different platforms. ponents neither deal with server-side sys- for their projects: • Applets: Applets are programmes tem-level programming nor with client- • Portability: There is only one version running via web browsers. Therefore, no side interfaces and therefore form a mid- of source code and one version of bina- installation is necessary on the side of the dle tier containing business logic only. ries for all operating system-architectures web client and also no maintenance. This gives much more flexibility when and hardware platforms. As a result of the Computations are performed on the client coping with changes of business rules. architecture-neutral capabilities, Java computer thereby off-loading our servers. • Lots of free software and a lively com- programmes execute on any platform for The next two bullet lists summarise munity with newsgroups • Object-oriented: This implies en- Table 1: Contact Points and Java Links in House. capsulation of data in objects and support of code-reuse. Service Contact Point • Security: Various security mechanisms prevent malicious code from be- Java Interest Group (JIG) Mailing List [email protected] ing executed. Send mail to Tim Canavan • Multi-threaded: Java supports thread [email protected] for registration. synchronisation and enables an appli- Java tools in archive and dmd Bob Kemp [email protected] cation to perform several tasks in paral- UNIX domains lel even if the underlying operating sys- Java tools for archive operations and Markus Dolensky [email protected] tem is not capable of multitasking. ecf domain • Late binding: Java is capable of dy- JSky Home Page http://archive.eso.org/JSky/ namically linking in new class libraries or JSky Mailing List [email protected] instance variables at runtime. Send mail to [email protected] • Robustness: Strict type checking, no for registration. pointers, no explicit memory (de)alloca- JSky Repository ftp://ftp.archive.eso.org/pub/jsky tion, enforcing boundary checking when P2PP Project Page http://www.eso.org/~amchavan/projects/j accessing arrays p2pp/development-docs.html Applet specific advantages • No installation: Just requires a Java Table 2: Applet Gallery. enabled web browser. • Automatic software update: The lat- Applet Name Description URL est software version is loaded whenever ESO ETC ESO Exposure Time Calculator http://www.eso.org/observing/etc/ an applet is invoked. There exists only one copy of the software on the web server. Isql JDBC interface to databases http://archive.eso.org:8009/ sample/gateway.html Weaknesses JIPA Download Page of Java http://archive.eso.org/java/jipa1.35/ • Rapid development led to a steady Image Preview Application stream of new Java versions and frame- JPlot Applet for Jitter Analysis http://archive.eso.org/archive/ work specifications rendering it a moving jplot.html target. It is only slowly settling down. Spectral HST Preview Applet for Spectra http://archive.eso.org/preview/ • Write once but install (in case of ap- preview/preview_hst/Y29E0305T/ plications) and test everywhere: Platform fits/spectral/4 specific bugs such as incompatibilities be- TimeGraph [10] Applet for System Load Monitoring http://www.hq.eso.org/bin/ tween JVMs are hard to detect and cir- tg?schema=sysload cumvent.

30 Figure 1: Alpha Release of STScI's archive browser StarView II. It uses ECF's utility JIPA to display a sky region from ESO's copy of the DSS catalogue together with an HST aperture overlay.

• Imposes strict adherence to object grammer can fine tune the look and feel • To provide a cleaner, more intuitive oriented design in contrast to C++. In fact, of the application to the desired level. user interface. One of the main aims of this could be seen as an advantage as DBB is used for the new Phase II Pro- the Swing graphics library for Java is to well. posal Preparation System (P2PP) V.2. provide platform-independent high-level • Programmer has no direct control graphical components. over system resources. Therefore flaws DocumentView: A way of linking Design Patterns with names like Fac- of built-in services like the garbage col- GUIs and Objects tory and Visitor [2] were used wherever lection algorithm are hard to circumvent. DocumentView is a set of classes that reasonable in order to reduce the learn- link object properties with GUI components. ing curve for future maintainers. P2PP V.2 Applet specific weaknesses For instance, one can link a text field can function as a 2- or 3-Tier client/serv- • Security restrictions make it difficult widget with the title of an Observation er system with a relational database back to implement basic features like saving Block (OB). Changes in each affect the end and an optional business object mid- the programme status and printing. other, and thus multiple GUI components dle tier. However, users can work com- • Lacking browser support, latency displaying the property are kept synchro- pletely off-line, saving unfinished work on problem due to limited bandwidth and the nised. DocumentView is a component their local machine. Connection to the lack of adequate caching mechanisms used for the P2PP tool described below. ESO OB repository is only needed in or- can lead to frustration of users. der to submit their OBs for execution. A New Generation Phase II Proposal Preparation Tool Java Interface to VLT Control 2. Application Briefs Software The successor to the current version DBB – The Database Browser of P2PP (Phase II Proposal Preparation This class library allows Java appli- Project System) is an application being devel- cations to interact with the non-Java VLT The Database Browser Project is oped using the Java programming lan- Control System. It provides interfaces to aimed at developing a library of Java guage. The new version is largely back- the the VLT Central Common Software. classes which can be used to create in- wards compatible with the current sys- These interfaces are implemented as teractive database browsing and editing tem, and its basic purpose is to provide Java Native Methods, relying on the lega- tools. Many data-centred applications help for the preparation of Observation cy VLT Common Software libraries im- share a common structure: A user wants Blocks (OBs) both at the observer's home plemented in C/C++. to select a set of items (rows) from a data- institution and at the telescope. The new library allows the develop- base according to some searching and The decision to rewrite version 2 of the ment of high-level interfaces and VLT sorting criteria, reviews a summary of the system in Java was based on two fun- control applications in Java. It was returned items, edits one or more items damental requirements: made available for the first time as part and saves the edited items back to the • To provide easier support for multi- of the OCT99 VLT Software Release. database. The DBB classes provide the ple platforms, since Java libraries are P2PP V.2 uses this interface to trans- application programmer with a set of tools available for most operating systems and mit OBs to the VLT Unit Telescopes in to quickly develop interactive tools of that are extremely consistent in their behav- Paranal and to get status information kind. By specialising the library, a pro- iour across platforms. about their execution.

31 a Bean Builder, ideally with little or no hand-written code necessary. The prototype can display FITS images and supports zooming, panning, colour maps, and automatic cut level setting. Initial tests indicate that the performance is satisfactory for small images, as long as the JAI native library (mlib) is available. At the time of writing, this is only the case under Windows and Solaris, although a Linux version is due out soon. The current version does not yet handle large images well, but can be optimised to do so. The FITS reader is based on Thomas McGlynn’s Java implementation, which currently always reads in the entire image. A future version will provide support for splitting the image into tiles, which is the only way to efficiently handle large images in JAI. The image display application (Fig. 2) currently has a hard-coded layout. In the end, the goal is to build the application from JavaBeans and allow the users to change the layout, add new panel items, remove unwanted ones, etc. The Java implementation of the catalogue browser looks pretty much like the Skycat version, so far, although this will likely change. The panel layout can be generated automatically, based on an eX- tensible Markup Language (XML) catalogue description and the catalogue data itself may also be in XML [9]. The new Figure 2: A first prototype of an image viewer has been developed, based on the JAI (Java Ad- catalogue interface should be more flex- vanced Imaging) package from Sun. ible than the Skycat version and allow various catalogue types to be supported by Java Calibration Database Manager The applet is based on Leigh Brook- means of a catalogue registry. shaw’s GNU plot library [4]. JSky is an open development platform The Calibration Database includes a and contributions of general-purpose or collection of observation data and re- ST-ECF Applets for the Archive astronomy-related software and compo- duction procedures representing as Interface nents are welcome. There is a JSky home completely as possible the observing ECF started its efforts in this field about page [Table 1] as well as a JSky ftp repos- modes and configurations of the differ- four years ago. Early on applets were re- itory [Table 1]. ent VLT instruments. The Calibration leased on the web. They are used for pre- viewing HST spectra and images [5]. Database Manager is the standard ap- References plication for browsing, editing, populating There is JPlot – a plot utility for jitter analysis of HST observations [6]. JPlot was and aligning the local calibration database [1] A. Walsh, J. Fronchkowiak, 1998, “Java contents. This Java application is the in- used for quality control purposes in first Bible”, IDG Books Worldwide Inc, p. 7–18. terface between the pipeline local cali- place and later on put on-line as a pub- [2] Gamma et al., 1995, “Design Patterns: El- bration database and the ESO archive lic archive service. There is also Skymap, ements of Reusable Object-Oriented system. It allows the ESO Quality Con- a simple applet that draws locations of Software”, Addison-Wesley. trol Scientists of the Data Flow Operations HST parallel observations on a map. [3] P. Ballester, A. Disaro, D. Dorigo, A. Modi- group to request and submit files to the As a logical next step, more ambitious gliani, J.A.Pizarro de la Iglesia, “The VLT archive, and to structure calibration projects followed which are currently ac- Data Quality Control System”, The Mes- information to the formats supported by complished in collaboration with other in- senger 96, p. 19. stitutions like STScI and ISO IDC [7], [8]. [4] GNU Graph Class Library the pipeline and needed for long-term ar- http://www.sci.usq.edu.au/staff/leighb/graph/ chival and trend analysis. The Calibration The aim is a pure Java interface to both [5] Dolensky et al., 1998, “HSTArchive Ser- Database Manager and other quality con- HST and ESO archives (Fig. 1) that will vices Implemented in Java”, The Mes- trol applications are described in more de- serve as a more powerful alternative to senger 93, 26. tail in a previous issue of The Messen- the current web interface. [6] Dolensky et al., 1998, “How the Analysis ger [3]. of HST Engineering Telemetry Supports JSky – Reusable Java the WFPC2 Association Project and En- ESO Exposure Time Calculator Components for Astronomy hances FOS Calibration Accuracy”, The The user interface of the on-line Ex- The JSky development effort aims to Messenger 93, 23. provide a library of reusable Java com- [7] Binegar et al., 1999, “Browsing astronom- posure Time Calculators provided for var- ical archives with StarView II and inter- ious VLT and NTT instrument as well as ponents for astronomy. Based on expe- acting with analysis tools such as JIPA”, the Wide-Field Imager was implement- riences with the Skycat application, the in ASP Conf. Ser., ADASS 99, in press. ed as a combination of HTML and Java. first components being developed are for [8] http://archive.eso.org/~mdolensk/publ/jskybof/ An observer specifies the relevant pa- image and catalogue display and ma- [9] http://vizier.u-strasbg.fr/doc/astroxml.htx rameters in a web form, the necessary nipulation. The goal is to eventually have [10] TimeGraph Download Page computation takes place on an ESO serv- a collection of general-purpose and as- http://www.ece.utexas.edu/~tbieleck/ er and the results are returned in a page tronomy-related JavaBeans that may be containing HTML text and a plot applet. easily assembled into applications using E-mail: [email protected]

32 EIS – THE ESO IMAGING SURVEY

The ESO Public Imaging Survey A. RENZINI and L. DA COSTA, ESO

1. Background mented by “privately” acquired data for South including the fields covered by specific applications, such as in the case the various HST cameras, and four In the era of 4-m-class telescopes it of second-epoch or narrow-band filter ob- adjacent pointings in the AXAF deep field. was common to select targets and/or servations. In both cases the multicolour UBVRI- prepare observations using deep Schmidt A Working Group (WG) was therefore JHK data should have been deep enough plates for which full sky coverage was appointed in 1996 to design the survey to provide accurate photometric red- available down to m ~ 23. Now, in the and to supervise the work of the Survey shifts and to select Lyman-break galax- blossoming era of 8-m-class telescopes Team charged to carry out the observa- ies. In practice, this required limiting one needs much deeper survey data tions, data reduction and distribution of magnitudes in the range 26–27 in the to feed high-throughput, high-multiplex survey products. Both the WG and the optical passbands and 23–24 in the spectrographs able to reach at least two Survey Team were assembled from the infrared. magnitudes deeper, or to find relatively community itself, with ESO providing the The one-year period (July 1997–July rare objects such as clusters of galaxies, coordination and the required facilities 1998) available to achieve the primary brown dwarfs, or trans-neptunian objects and resources. goals of EIS-WIDE represented a formi- (for an extensive list of such potential tar- The main objectives of the resulting dable task, as no consolidated experi- gets see Renzini 1998, The Messenger, ESO Imaging Survey (EIS) were to ence existed at ESO or in the communi- 91, 4). To a large extent, the scien- conduct a public optical-IR imaging ty. Apart from the limitations due to the tific outcome of an 8-m-class telescope survey at the NTT, to reduce the data, weather, the goals were only met thanks critically depends on the availability of construct object catalogues, and select to the enthusiastic response of many sci- preparatory imaging surveys, and the de- from them special classes of objects entists from the community who brought mand for deep and wide surveys will in- of potential interest for VLT pro- their specific experience to ESO and to tensify as other large-aperture tele- grammes. All this within a short time the resources made available at ESO to scopes become operational world-wide scale, before the start of operation of the support the temporary relocation of and the competition becomes more and first Unit Telescope of the VLT. In addi- these scientists. The contribution of more fierce. tion, EIS was seen as a first step towards ESO and ECF staff was also critical for CCD and computer technology have developing an appropriate environment several aspects of data processing, evolved at a rapid pace over the last two for carrying out future surveys, pub- archiving, and distribution. decades, and multi-colour, digital imag- lic as well as private. Indeed, the software With the advent of the Wide Field Im- ing surveys of appropriate area and depth tools developed by the programme were ager (WFI) at the 2.2-m telescope are within reach. Even though there are regarded as an integral part of the (Baade et al 1999, The Messenger, 95, a number of ongoing optical/infrared dig- survey products, to be disseminated 15) a “WFI Pilot Survey” was designed ital all-sky surveys (DENIS, SDSS, throughout the ESO community to facil- by the WG and recommended by the 2MASS), none of them are particularly itate the reduction and analysis of wide- OPC to be conducted by the EIS Team. suitable for feeding targets for large tel- area imaging data. The dissemination of The scientific goal of the Pilot Survey was escopes (for most possible targets they such tools was seen as essential for the essentially to complete that part of the are not deep enough) nor will they be community to take full advantage of ded- EIS-WIDE project that could not be done publicly available in the near future. Sur- icated imaging telescopes such as the due to bad weather during the allocated veys more suitable to large telescopes MPG/ESO 2.2-m telescope, now already nights (it is worth recalling the impact of are being conducted at several obser- in operation, and the VLT Survey Tele- El Niño in 1997–98). The “pilot” desig- vatories (e.g., CFHT, CTIO, La Palma, Kitt scope (VST, Arnaboldi et al. 1998, The nation was meant to reflect the devel- Peak) with the products becoming pub- Messenger, 93, 30) to be installed on opment period required for the major up- lic within their respective communities. Paranal. grade of the survey software to enable it Groups with specific science goals are to pass from one-CCD (EMMI) or two- also conducting ambitious surveys, but 2. The EIS Project and the WFI CCD camera data to the eight 2k × 4k again these data are unlikely to become Pilot Survey CCDs of WFI. accessible to general ESO users in a The companion article by da Costa et short-time scale. In practice, to combine as far as posal. in this issue of The Messenger de- To meet this challenge, in 1996, ESO sible wide-area coverage and depth, EIS scribes in some detail what was achieved took the initiative of proposing to car- consisted of two complementary projects: by these surveys, with emphasis on the ry out a public survey to fulfill at least EIS-WIDE and EIS-DEEP. The goal of survey products. All the data from EIS- some of the short-term needs of the EIS-WIDE was to cover with EMMI@NTT WIDE, EIS-DEEP and the WFI Pilot Sur- ESO community in view of the first sci- four patches in the sky, 6 square degrees vey were delivered to the community in entific runs of the VLT. Without hamper- each, providing V and I data to a limiting due time. Over the last three Calls for Pro- ing the possibility of groups to conduct magnitude of VAB ~ 24.5 and IAB ~ 24. posals (P63–P65), over 100 proposals for their own surveys, a public survey offers This part of the survey was optimised to four different Panels, were based on the a number of advantages. In particular, detect candidate clusters of galaxies at EIS data, or referred to EIS. Follow-up ob- data can be used for a broad range moderate to high redshift (z <~ 1). A ~ 2- servations were conducted using ESO of scientific applications, thus result- square-degree fraction of one of the telescopes as well as other facilities to ing in a substantial saving of telescope patches was also to be observed in U and prune samples in preparation of VLT pro- time (that can be as high as a factor of B to allow for the identification of QSOs, posals. Targets selected from the EIS 10!), compared to the case in which metal-poor stars, etc. data were also used during Science Ver- each individual science group has to se- The goals of EIS-DEEP were to cov- ification of UT1/Antu as well as of cure its own survey data. Moreover, pub- er with SUSI2 and SOFI three adjacent FORS1 Commissioning and Science Ver- lic survey data can be easily comple- pointings in the Hubble Deep Field ification.

33 3. The Public Survey, Period 64 myth platform of the same telescope (see thus require more information to be prop- to 67 Renzini 1999, The Messenger, 96, 13). erly characterised, relative to radio or The accuracy of fibre positioning, which space-based data. The usefulness of Public Surveys in requires accurate target coordinates to The EIS/Pilot Survey project made support of VLT programmes being start with, is critical. Existing stellar cat- available four kinds of data products: as- generally recognised, the WG at its alogues fail to give the information that trometrically and photometrically cali- meeting in September 1998 recom- is necessary to operate FLAMES, i.e., co- brated images, filtered and verified ob- mended ESO to issue a “Call for Ideas” ordinates with the required astrometric ject catalogues drawn from single-colour for future public surveys to be con- accuracy <~ 0″. 1 rms) combined with mul- coadded frames, derived catalogues ducted with the 2.2-m and NTT tele- ticolour information. The tentatively se- such as colour catalogues, and target scopes (The Messenger 94, 31). The re- lected fields for the Pre-FLAMES Survey lists. Each release was also accom- sulting “ideas” were discussed in depth are listed on the EIS web pages. Inter- panied by papers describing the ob- by the WG in March 1999, and two opti- ested scientists are welcome to give sug- servations, data reduction, the different mised proposals emerged. A small short- gestions to complement and/or improve types of catalogues available and the tar- term proposal to complete the EIS-WIDE the list. get lists produced. The object catalogues survey reaching the goals as originally The Public Survey proposal was sub- were also essential to validate the data stated, and a two-year Public Survey mitted as a Large Programme in re- by computing simple statistics such as the programme. The latter consists of three sponse to the Call for Proposals for Pe- number counts of stars and galaxies, parts: (1) a deep multi-band (UBVRI) riod 64, applying for a total of 81 nights: colour-colour diagrams for point-sources optical survey with the WFI over three for the DPS 54 nights at the 2.2-m tele- and the two-point angular correlation fields of one square degree each; (2) a scope and 15 nights at NTT for its function for galaxies. This allowed the deep near IR survey with SOFI, cover- IR part, plus 12 nights at the 2.2-m photometric zero-points, star/galaxy clas- ing two 450 square arcmin fields select- telescope for the Pre-FLAMES Survey. sification, and uniformity of the data to be ed within the area observed in the opti- This Large Programme was proposed tested. cal; and (3) a shallower set of WFI ob- for a two-year period, hence being dis- The primary reason for adopting this servations in V and I of selected stellar tributed over ESO Periods 64 through data release model was the strong de- fields in preparation for FLAMES at 67. Details of the Public Survey, includ- sire to have verified data publicly avail- UT2/Kueyen (hereafter the “Pre-FLAMES ing the original proposal, can be found at able in the shortest possible time to meet Survey”). http://http.hq.eso.org/science/eis/ the stringent deadlines imposed by the The selected fields for the deep op- public-surveys/ imminent start of the VLT regular opera- tical/IR survey (hereafter Deep Public This Public Survey “large proposal” tions, which began with the submission Survey, DPS) are given in Table 1. The was approved and recommended by the of the proposals for Period 63 (deadline main scientific drivers for this part of OPC, and 18 nights with [email protected] and October 1, 1998). However, this model the Public Survey are the search for 8 nights with SOFI@NTT have been is not the most suitable for long-term proj- high-redshift objects: galaxies (e.g. Ly- scheduled in Period 64. To increase the ects such as the Public Survey described man-break galaxies), clusters and QSOs, participation of the community in the Pub- in the previous section. The time base- as well as the possibility to de- lic Survey, an Announcement of Oppor- line for the completion of the project is rive photometric redshifts from such a tunity was also issued for groups in the much longer, and intermediate products large, homogeneous, multicolour data- community to carry out the observations may be valuable for different applications. base. The goals and requirements in and data reduction of the infrared part of While it is obviously highly desirable to terms of limiting magnitudes are sim- the DPS, following the same guidelines cut to the minimum the time lag between ilar to those of EIS-DEEP described of the EIS Team. Following an examina- the collection of the survey data and above except that DPS is meant to cov- tion of the proposals, the WG has already the VLT follow-up, the sheer volume of er a much larger area. The same data- selected the Team that will be in charge the data expected from [email protected] re- base will also provide a unique and ur- of this part of the survey (PI H. Jørgensen, quires more time for detailed verifi- gently needed tool for studies of Galac- Copenhagen). cation of the data products. In addition, tic structure and stellar populations. The The Public Survey in periods 64–67 will users may have their own ideas on how DPS is therefore designed to provide tar- provide the ESO community with a com- to best extract catalogues for their spe- gets for several of the first-generation VLT petitive edge in the realm of the high-red- cific scientific purposes, while final re- instruments, with emphasis on FORS1 shift universe as well as in stellar stud- sults for a given sky area can only be pro- and FORS2, ISAAC, UVES, VIMOS, and ies. Furthermore, it will allow the contin- duced after all the observations are com- NIRMOS. uation of the development effort of EIS pleted. On the other hand, the Pre-FLAMES that will guarantee a smooth transition be- One solution to accommodate these Survey is of crucial importance for stel- tween the [email protected] and the VST that will various constraints is to provide both lar astronomy studies with the FLAMES be equipped with ΩCam, four times larg- a prompt release, soon after an ob- System on UT2/Kueyen, where the er than WFI in both field of view and num- servation run is completed, and a OzPoz fibre positioner can pick up ber of pixels. more complete release on a time table ~ 130 targets over a field of 25 diameter synchronised with the ESO observing (well matched by the field of view of WFI) periods. In this framework one pro- and feed the MEDUSA mode of GI- 4. The Public Survey Products vides rapid access to reduced – but RAFFE. Alternatively, 8 fibres from and Distribution still unverified – data, thus allowing OzPoz could send the light of as many users to extract objects directly from targets to UVES, over to the other Nas- Among the many requirements of a the images. Meanwhile, the Survey Public Survey, the most important is per- Team will work to prepare complete haps that the data products be easily re- and fully verified data releases. The aim trievable by external users. While this is to strike a balance both between the Table 1: Deep Public Survey Selected may seem trivial, practice has shown that needs of the community and the opera- Fields. it is not always simple to reconcile the tional requirements, and between speed needs of different users – with the implied and quality. Following this approach, Field α(J2000.0) δ(J2000.0) large variety and size of the products – the distribution of data accumulated with other practical constraints such as from the Public Survey will consist of Deep 1 22 43 00 –39 58 00 efficiency and resources. In addition, data prompt and final releases, which will oc- Deep 2 03 32 28 –27 48 00 from ground-based optical observations cur on different time scales, as described Deep 3 11 22 00 –21 35 00 are subject to atmospheric conditions and below.

34 4.1. Prompt releases the quality of the data, the data reduction compressed images, which can lead to and verification procedures, and the de- a significant reduction (as much as a fac- The prompt releases will consist of scription of the content of the derived cat- tor of 500) in the size of the images. How- products that are generated directly by alogues and of the methodology used ever, such compression schemes in- the EIS-WFI pipeline. These will provide in their preparation. These papers are evitably involve some loss of information the opportunity for users to have a “quick- also important as a reference, thereby and the scientific value of the resulting look” at the data soon after they have recognising the commitment of ESO in products depends on the specific appli- been acquired. These releases will con- producing the data and in making them cation. Experiments conducted by the sist of: available to the ESO community. While Survey Team indicate that for astromet- • Astrometrically and photometrically no direct scientific exploitation of the data ric applications very little is lost, even for calibrated coadded images of dithered, will be done in these papers, they would very large compression factors. In fact, single-pointing exposures and associat- still represent a reasonable reward for the the loss in accuracy is <~ 0.5 of pixel size, ed weight maps. The astrometric and members of the Survey Team for their ef- comparable to the estimated internal er- photometric accuracy shall be better than fort in conducting such a survey of pub- ror of the astrometric solution. Examples 0.″05 and 0.1 mag, lic utility. of compressed images have been made • Catalogues of objects extracted from available in the most recent data release. these images using SExtractor, in the 5. Data Distribution Logistics With the advent of the WFI it seems that form of FITS binary tables. the combination of compressed images Such releases are expected to start A daunting task for public wide-field for large fields and full resolution cut-outs as soon as the ongoing work of up- imaging surveys is finding the ways for small regions should be seriously con- grading the EIS-WFI pipeline is com- and means of exporting the data to out- sidered. pleted and the new version properly test- side users. To give an indication of the In summary, to a large extent the scope ed, which should occur by the end of scale of the problem it suffices to men- of the Survey Team work and the range 1999. tion that the Pilot Survey alone has pro- of data products have been dictated by These data together with SExtractor, duced over 0.5 Terrabytes of data in the recognition of the difficulties in ex- LDAC and Drizzle softwares, which are roughly 11 nights of WFI observing, while porting full-resolution images, as result- publicly available and can be retrieved 66 nights have been assigned for the ing from the direct experience with the from the EIS home page, or directly from Public Survey at the [email protected]. Even af- EIS project, and the expected trend with their respective authors, will enable ter co-addition of all dithered images, a the [email protected] and the ΩCam at the VLT users to work in parallel to the Survey single-passband 6-square-degree mo- Survey Telescope. One needs to effi- Team, starting from the same basic data saic consists of 6.5 Gigabytes, and a ciently translate terabytes of imaging data and at the same time. Users will, there- complete data set for the recently ob- into megabytes of useful information in fore, be able to evaluate the data and pre- served EIS patches (combining data from the form of the various data products pare, if they so desire, their own cus- EMMI and WFI) comprises 64 Gb of im- made available by the Survey Team (sin- tomised catalogues without having to rely ages and weight maps, and 32 Gb of con- gle and multi-band object catalogues, tar- solely on those derived by the Survey text maps. The stellar fields comprise an- get lists, cut-outs, postage stamps). In this Team. Moreover, new tools are in the other 63 Gb of images. Therefore, with context it is important to emphasise once process of being installed which will fur- the presently available technology it is un- more that the primary goal of the ESO ther enhance the scope of work that can realistic to consider the distribution of pix- public surveys has been to support VLT be done outside ESO without the need el maps for an undetermined number of observations and not to be an end in it- for distributing large amounts of raw or requests. To make the distribution pos- self. processed data. sible, with the resources currently available at ESO for this task, a certain num- 6. Future Plans 4.2. Complete releases ber of constraints are unavoidable. The distribution of bulk data must be limited The Survey Team and the ESO Complementing these prompt re- to well-defined packages of reduced data, Archive Group continue to explore oth- leases, complete releases of verified produced semi-automatically; distribution er ways to provide external users with data including all the data accumulated of large volumes of data must be limited more flexibility in their use of the sur- during all previous observing periods to certain media, e.g. DLT model 7000 vey data. In the short term it will be will occur every 6 months, namely on which store 35 Gb and are produced in possible to use the SKYCAT interface February 15 for odd periods and August about 2 hours (5 Mb/s). Large volumes to run SExtractor on image cut-outs 15 for even periods. These will include of data cannot be requested by a simple extracted from WFI mosaics, of suf- not only images but also a host of de- click on the web page, but will have to be ficient size (12′ × 12′) to include the rived products which require more time requested by submitting a statement of field-of-view of the FORSes and of infor their proper preparation and verifi- purpose which should include a scientif- dividual VIMOS and NIRMOS cameras, cation. These releases will include: ic justification and demonstrate that re- regardless of their respective orienta- • Tools to allow the extraction of image sources are available at the home insti- tions on the sky. In the long term, efforts sections (cut-outs) from image mosaics tute to handle the requested amount of will be made to implement an object- and associated weight maps data. oriented database which will allow users • Co-added, mosaiced images of in- Note that even a single WFI frame (267 to fully explore the multi-dimensional po- dividual patches and associated weight Mb) cannot be transferred by FTP in a sition, magnitude and colour space de- maps reasonable time, under normal condi- fined by the objects extracted from the im- • Colour catalogues for multi-passband tions. Major improvements will have to aging surveys. The development of observations await the development of new storage such a database, where astronomical ob- with all image products being photo- media and of broadband technology. This jects extracted from survey data can be metrically and astrometrically calibrated also raises the issue of how rapidly and stored, updated and associated with all as specified above. widely new storage media become avail- the information available from internal and Each release will be followed by an able throughout European institutes. By external sources, and of tools for search- update of the web and will be ac- now it should be clear that the rate at ing special categories of objects is an es- companied by papers prepared by the which images can currently be dissemi- sential element for the efficient use of Survey Team. These papers are an in- nated lags far behind the rate at which public survey data, as well as of all oth- tegral part of the data distribution as they are acquired. er imaging data that will also become they describe in detail the observations, In order to overcome this problem, the publicly available after the proprietary community is urged to consider using period.

35 7. Conclusion over the past two years as well as short- The new policy outlined here is meant term visits of leading experts acting as to bring the data promptly to the users, Besides the data, arguably one of the consultants. While about half of the long- while allowing the Survey Team more most important contributions of the EIS term visitors have already returned to time to deliver the final products. Budget project has been to foster a collaboration their home institutes, a large fraction con- constraints obviously limit the size of the between different European groups to de- tinues to contribute to EIS in different Survey Team, which cannot grow at the velop a comprehensive imaging data ways. same pace as the data flow. Hopefully, analysis pipeline for the timely and full As stated over two years ago, EIS was this new model will stimulate new uses exploitation of the data that will become conceived as an experimental project not and checks of the data. Experience has available from wide-field imagers. EIS has only to provide targets for the first year shown that, regardless of the care tak- both benefited from and contributed to of VLT (see da Costa et al. in this issue en, handling large volumes of data is a this ongoing effort, by capitalising on the of The Messenger) but also to explore the difficult task and only by the exhaustive available expertise and by providing data feasibility of conducting public surveys us- use of the data in different applications and a realistic framework and time-table ing dedicated wide-field imaging tele- can problems be detected. Therefore, the to test concepts, designs and imple- scopes. Indeed, with its 9 × 9 arcmin field participation and the feedback from the mentations of pipelines, database archi- of view, EMMI@NTT was far from being users is essential to ensure the quality of tecture and data archiving and distribu- the ideal instrument for wide area sur- the data. Users are also encouraged to tion schemes. Since the first feasibility veys. Yet, without the experience gained periodically visit the EIS home page to study carried out in 1995, over 30 peo- with EIS it would now be much more dif- monitor any new version of the various ple have contributed to the EIS effort ficult to efficiently deal with the [email protected] releases, and report any possible error. among EIS visitors, WG members and data, and in the medium term perspec- ESO staff and fellows. tive with the almost 10 times larger data 8. Acknowledgements EIS has also provided a stimulating en- flow from the ΩCam@VST. EIS has vironment for the introduction of students served to prototype and build a frame- We would like to thank the EIS Team to modern techniques of data reduction work for this type of long-range pro- and the ESO Archive Group for their con- and analysis and the means to dissem- gramme, essential for the whole ESO tinuing efforts to explore ways of im- inate their experience throughout the community to benefit from such tele- proving the accessibility of the data by ex- community. This has been achieved by scopes, and to do so to the largest pos- ternal users, as well as the NTT and 2.2- the EIS Visitor Programme which has sible extent independently of the effec- m Telescope Teams for their cooperation. sponsored long-term visits, ranging from tive resources available at the individual 6 months to over 2 years, of 15 people institutes. E-mail: [email protected]

ESO IMAGING SURVEY: Past Activities and Future Prospects L. DA COSTA1, S. ARNOUTS1, C. BENOIST1, E. DEUL1,2, R. HOOK3, Y. -S. KIM1, M. NONINO1,4, E. PANCINO1,5, R. RENGELINK1,2, R. SLIJKHUIS1, A. WICENEC1, S. ZAGGIA1

1European Southern Observatory, Garching b. München, Germany 2Leiden Observatory, Leiden, The Netherlands 3Space Telescope – European Coordinating Facility, Garching b. München, Germany 4Osservatorio Astronomico di Trieste, Italy 5Dipartimento di Astronomia, Universtà di Padova, Italy

1. Introduction ed on the MPG/ESO 2.2-m telescope at for a total integration time of 300 sec. La Silla. The observations were carried out i n The ESO Imaging Survey (EIS) proj- The purpose of this contribution is to the period July 97–March 98 and cov- ect is an ongoing effort to carry out pub- briefly review the results of the original ered four patches of the sky south lic imaging surveys in support of VLT pro- EIS and to give an update of the results of δ = –20° and in the right ascension grammes. Background information on the obtained from the observations carried range 22h < α < 10h, thus producing tar- original and future goals of the pro- out as part of the Pilot Survey. The on- gets suitable for the VLT almost year- gramme can be found in Renzini & da going work to develop an advanced round. The locations of these fields are Costa (1997) and in a companion article pipeline for handling data from large CCD shown in Figure 1 and their approxi- by Renzini & da Costa in this issue of The mosaics and of facilities to make access mate centres are given in Table 1, which Messenger. to the data products easier to external also lists the area covered in each The first phase of the project, which users are also discussed. passband. The typical depth of the sur- started in July 1997, consisted of a mod- vey, as measured from the co-added mo- erately deep, large-area survey (EIS- saics and with the magnitudes ex- WIDE) and a deep optical/infrared sur- 2. EIS-WIDE pressed in the AB system, is given in vey (EIS-DEEP), with the observations Table 2 which lists: in column (1) the pass- being conducted at the NTT. EIS has This part of the survey consisted band; in column (2) the median seeing; recently reached another milestone with of a mosaic of overlapping EMMI-NTT in column (3) the 1σ limiting isophote; the completion of a Pilot Survey us- frames (9 × 9 arcmin) with each po- in column (4) the 80% completeness ing the Wide-Field Imager (WFI) mount- sition on the sky being sampled twice magnitude (e.g., Nonino et al. 1999); in

36 of the other patches, giving priority to the I-band survey. Despite the smaller total area and the varying quality of the data, it was possible to meet most of the primary science goals of the survey. Particularly successful has been the I-band survey covering ~ 17 square degrees to a limiting magnitude IAB ~ 23.5, currently the largest survey of its kind in the southern hemisphere. In addition the survey provided ~ 3 square degrees in V I and ~ 1.5 square degrees in B V I. From these data, samples of distant clusters of galaxies and quasar candidates were compiled (Olsen et al. l999a,b, Scodeggio et al. 1999, Zaggia et al. 1999) and used in several follow-up observations by different teams. The data, comprising astrometrically and photometrically calibrated pixel maps, derived catalogues and target lists, were made publicly available in March and July 1998 (Nonino et al. 1999, Prandoni et al. 1999, Benoist et al. 1999), Figure 1: Sky projection showing the location of the original EIS-WIDE patches (red), the EIS- meeting the planned deadlines. To ease DEEP AXAF and HDF-S fields (blue), and the stellar fields of the Pilot Survey (orange). Also the access to the data, a web interface in blue are the new fields selected for the Deep Public Survey to be carried out in periods 64–67. was created to enable external users to Note that one of the regions selected coincides with patch A and the other with the EIS-DEEP request survey products, to extract image AXAF field (see Fig. 2). The third field (Deep-3) has been selected in a region with lower ab- cutouts and to examine the selected tar- sorption and less crowding than patch D. The right ascension range was chosen to distribute gets on-line. as much as possible the observations throughout the year. The accuracy of the relative astrometry is ~ 0.03 arcsec, making the coordinates sufficiently accurate for the preparation of multi-object spectroscop- Table 1: EIS-WIDE: Sky Coverage (square degrees). ic observations. The internal accuracy of the photometric zero-point has been es- α δ Patch 2000 2000 B V I timated to be better than 0.1 mag, consistent with the results obtained from A 22 43 03 –39 58 30 – 1.2 3.2 comparisons with other data (e.g., DE- B 00 48 22 –29 31 48 1.5 1.5 1.6 NIS). However, as more data in different C 05 38 24 –23 51 54 – – 6.0 passbands became available, gradients D 09 51 40 –21 00 00 – – 6.0 of the photometric zero-point, on a degree scale, have been noticed and are cur- Total 1.5 2.7 16.8 rently being investigated. The results of these investigations will be reported in due time. column (5) the 3σ limiting magnitude reached, not all the objectives were 3. EIS-DEEP within an aperture twice the median met. Among the most important omis- FWHM; and in column (6) the originally sions were the observations of a com- This part of the survey consisted of proposed magnitude limits (Renzini & parable area in V and of patch B in the deep optical/infrared observations using da Costa 1997) expressed as in column U-band. The incompleteness was due to SUSI2 and SOFI at the NTT of the HDF- (5). Note that these numbers are rep- a combination of poor weather conditions S and AXAF fields shown in Figure 1. The resentative values since, as pointed out and unexpected overheads in the ob- original observations were carried out in the original papers (Nonino et al. servations. A total of about 42 nights were in the period August–November 1998, 1999, Prandoni et al. 1999, Benoist et al. assigned to this part of the project out of with all the data being publicly released 1999), the observing conditions varied which about 30% were lost due to the December 10–12, 1998 (da Costa et al. significantly during the course of the ob- weather alone, compromising especial- 1999b, Rengelink et al. 1999). The re- servations. Comparison of Tables 1 and ly the coverage of EIS patches A and B. lease included fully processed and cali- 2 to the original goals of the survey At the time, the Working Group (WG) in brated images, single-passband and (Renzini & da Costa 1997, and revised charge of overseeing the survey decid- colour catalogues, and a tentative list of in da Costa 1997) shows that, while ed to postpone the U-band observations U -and B dropouts. The centres of the the desired limiting magnitudes were of patch B and the V-band observations different fields observed with SUSI2 and SOFI, chosen to provide some overlap, are given in Table 3. The data Table 2: EIS-WIDE: Limiting Magnitudes (AB system). available for these fields are summarised in Tables 4 and 5 which pro-

Filter FWHM µlim 80% 3σ 3σ (proposed) vide, for each field, the following infor- (arcsec) (mag arcsec–2) (2 × FWHM) (2 × FWHM) mation: in column (1) the passbands available; in column (2) the total inte- B 1.0 26.5 24.6 24.2 24.2 gration time; in column (3) the number of V 0.8 26.2 24.4 24.0 24.5 frames; in column (4) the final FWHM as I 0.8 25.5 23.7 23.7 24.0 measured on the co-added image; in column (5) the 1σ limiting isophote; in col-

37 Figure 2: Schematic view of the various data sets available for the HDF-S and AXAF fields from optical/infrared observations conducted with SUSI2 and SOFI at NTT. The different colours denote the completeness of the multi-colour data.

Table 3: EIS-DEEP: SUSI2 and SOFI Pointings (J2000.0). umn (6) the limiting magnitude of a 3σ detection within an aperture twice the α δ α δ Field SUSI2 SUSI2 SOFI SOFI FWHM. In Table 4 the last column lists the originally proposed limiting magni- HDF1 22 33 29 –60 33 50 22 33 32 –60 33 30 tudes for EIS-DEEP, expressed as those HDF2 22 32 42 –60 33 50 22 33 00 –60 33 30 in column (7), assuming a 1 arcsec see- HDF3 – – 22 33 00 –60 37 59 ing (da Costa & Renzini 1998). To help visualise the fields and the different sets of data available, Figure 2 shows a AXAF1 03 32 16 –27 46 00 03 32 17 –27 46 10 schematic view of the EIS-DEEP obser- AXAF2 03 32 38 –27 46 00 03 32 37 –27 46 10 vations of the HDF-S and AXAF fields, AXAF3 – – 03 32 17 –27 50 35 with the colours indicating the com- AXAF4 – – 03 32 37 –27 50 35 pleteness of the multi-colour data. In the case of HDF-S the location of the WFPC2, STIS and NIC3 HST fields and their relation to the location of the EIS- DEEP fields are also shown. Table 4: EIS-DEEP: HDF-S Observations (magnitudes in the AB system). The highest priority was given to the observations of the HDF2 (WFPC2) field Filter t Nƒ FWHM µ 3σ 3σ (proposed) total lim for which data in eight passbands are (sec) (arcsec) (mag arcsec–2) (2 × FWHM) (2 × FWHM) available, by and large reaching the required magnitude limits, with the notable HDF1 exception of the I-band. Note that in this field the J and Ks observations reach half- J 10800 180 1.37 25.7 23.6 24.2 magnitude fainter limits than those in the H 3600 60 0.91 25.1 23.4 23.7 other fields, as suggested by the WG (da Ks 10800 180 0.90 24.9 23.1 23.1 Costa & Renzini 1998). Even though it was not possible to complete the obser- HDF2 vations in all the passbands for all the fields, the data gathered provide ~ 150 U 17800 22 1.00 28.2 26.3 26.6 square arcmin in J Ks and ~ 80 square B 6600 22 0.84 28.2 26.5 25.7 arcmin in UBVRI. When combined, the V 12250 49 1.27 28.3 26.5 25.8 optical/infrared deep observations pro- R 5500 22 1.05 27.3 25.4 26.0 vide a coverage of 15 and 25 square arcmin in eight and seven passbands, re- I 8800 44 1.11 27.0 25.0 26.2 spectively. These numbers include the re- J 10800 180 0.90 26.2 24.4 24.7 cently released data for the HDF-S STIS H 7200 120 0.85 24.9 23.2 24.2 and NICMOS fields from observations Ks 18000 300 0.96 25.5 23.7 23.6 conducted in July 1999 (see EIS home page). Unfortunately, due to bad weath- HDF3 er, no new optical data were obtained during the four half nights scheduled in July J 5820 97 1.18 25.5 23.5 24.2 1999 at the NTT, and the RI images of Ks 7440 124 0.86 24.9 23.2 23.1 HDF1, taken in 1998, proved to be un- suitable due to poor seeing and stray

38 light. The incompleteness of the EIS- a total of 8.5 nights in the months of April, Table 6: WFI Pilot Survey: Sky Coverage DEEP data relative to the original goals May and September. The observations (square degrees). is primarily due to the optical observations consisted of five 4-minute dithered ex- which were affected by both poor weath- posures, which allowed the coverage of Patch U B V I er and a variety of technical problems in 0.5 square degrees/hour in a single pass- 1998. An attempt to complete the optical band including overheads. A total of 18 A – 1.0 1.2 3.2 coverage of the HDF1 (STIS) field will be square degrees were observed, cover- B 1.25 1.5 1.5 1.6 made in June–July 2000, while the AXAF ing patches C and D in B and BV, re- C – 6.0 6.0 6.0 field has been chosen to be one of the spectively. The B-filter was used first be- D – 6.0 6.0 6.0 three fields selected for the new Deep cause it was the only broad-band filter Public Survey as shown in Figure 1 (see available during commissioning. In mid- Renzini & da Costa 1999). September new observations of Patch- Total 1.25 14.5 14.7 16.8 es A and B were conducted in V and U but the data have not 4. [email protected] Pilot Survey yet been reduced at the time of writing. Table 7: WFI Pilot Survey: Limiting Magnitudes (AB system). With the announcement of the com- Table 6 shows the cur- σ missioning of the [email protected], the WG rent status of the EIS- Band FWHM µlim 80% 3 recommended the undertaking of a pilot WIDE patches after (arcsec) (mag arcsec–2) (2 × FWHM) multi-colour survey over the area already combining the data covered in I – band, including the miss- available from EMMI B 0.9 26.8 24.5 25.0 ing U-band observations of patch B. The at the NTT with those V 1.0 26.4 24.1 24.6 motivation was to complete the original from the Pilot Survey, goals of EIS-WIDE and to steer the up- including the Septem- grade of the pipeline to cope with CCD- ber 1999 run. These data together with reached in B and V from the data gath- mosaics and the corresponding order of the 17 hours scheduled in service mode ered so far using the WFI are given in magnitude increase in the volume of data. for Period 64 should allow the com- Table 7. Even though originally proposed to be pletion of about 1.5 square degrees in Some of the scheduled time was un- carried out in dark time over the observ- UBVI and 15 square degrees in BVI. suitable for the observation of the patch- ing Periods 62 and 63, some adjustments These data should greatly broaden the es, either because the EIS-WIDE fields had to be made to accommodate the WFI range of applications that can bene- commissioning period. In the end, a to- fit from the EIS public survey to produce tal of 30 hours was used for the Pilot samples for follow-up observations with Survey observations during the com- the VLT. Preliminary estimates of the missioning period in January 1999, and median seeing and limiting magnitudes

Table 5: EIS:DEEP: AXAF Observations (magnitudes in the AB system).

Filter ttotal Nƒ FWHM µlim 3σ (sec) (arcsec) (mag arcsec–2) (2 × FWHM)

AXAF1

U 17000 21 0.90 28.0 26.1 B 6600 22 1.10 28.0 26.4 V 5500 22 0.88 27.7 25.7 R 5500 22 0.89 27.5 25.9 I 12600 21 1.31 27.0 24.9 J 10800 180 0.99 25.7 23.8 Ks 10800 180 0.95 24.9 23.2

AXAF2

U 13000 16 0.90 27.8 26.0 B 5400 18 1.10 27.9 25.8 V 5500 22 0.88 27.7 26.1 R 5500 22 0.89 27.5 25.7 J 10800 180 0.90 25.8 24.1 Ks 10800 180 0.90 24.8 23.1

AXAF3 Figure 3: Vector representation of the pattern J 10800 180 0.90 25.4 23.7 of the PSF anisotropy and the amplitude of its components for each chip of the CCD Ks 10800 180 1.00 24.7 22.9 mosaic, as measured on a V-band image obtained from a 20-min exposure. AXAF4

J 10800 180 0.70 25.6 24.1 Figure 4: Colour image of a section of EIS Ks 7200 120 1.00 24.4 22.6 patch D combining BV images of WFI with I-images of EMMI. ̈

Figure 5: A 16.1 × 9.4 arcmin section of the co-added I-band image, rebinned to a scale of 0.5 arcsec/pixel, of the cluster NGC 6558 located in Baade's window. The regions devoid of objects are associated with the interchip gaps. were not visible or because the nights cluded Baade’s window, globular (GC) upgraded EIS-WFI pipeline, and publicly were too bright. In these periods, a set and open clusters (OC). A total of 22 fields released on September 9, 1999. of selected stellar fields were observed (5.5 square degrees) given in Table 8 Preliminary results show that the per- as a test case for future preparatory pro- were observed in different passbands. A formance of the [email protected] exceeds ex- grammes for GIRAFFE and UVES using full description of the available data can pectations and demonstrate its compet- the fibre-positioner FLAMES, thus an- be found on the web. itiveness in the area of wide-field imag- ticipating part of the Public Survey ob- Altogether the observations from Jan- ing surveys. Analysis of the images servations now planned for P64–67, the uary to April 1999 yielded 400 Gby of raw shows that the point-spread function is so-called Pre-Flames Survey recom- data at a rate of about 40 Gby/night (0.5 very uniform across the entire field of mended by the WG in their meeting in Mb/sec), including calibration frames view and that PSF distortions are small March 1999 (see Renzini & da Costa (standard stars, dome and sky flats). All ~ 2%. This is illustrated in Figure 3 1999). The selected fields (Figure 1) in- the data have been processed, using the where the spatial distribution of the po-

42 larisation vector, representing the amplitude and direction of the PSF distortions, and the distribution of the amplitude of its two components are shown for each CCD chip of the WFI camera. To illustrate the results obtained by the pipeline, Figure 4 shows a colour composite of a section of patch D (23 × 19 arcmin) constructed from the combination of WFI B and V mosaics with the EMMI I-band mosaic. Since one of the primary goals of Pilot Survey is to complement the I-band observations, the co-addition of the WFI images was carried out on the EMMI pixel scale, about 10% larger than that of the WFI. By dithering the WFI exposures and by coadding the available frames using the weight maps, the final coadded mosaic shows no noticeable evidence of the inter-chip gaps, with the least sampled regions having an exposure time about 60% of that obtained in the best sampled regions. Another example can be found in Figure 5, where the co-added B-band image of the globular cluster NGC 6558, observed with 0.6 arcsec seeing, is shown. Note that in this case only two images have been co-added and the regions devoid of stars are the remains of the interchip gaps. The co-addition is carried out after Figure 6: Distribution of the differences in the coordinates of 700,000 pairs of stars in patch D the astrometric calibration of each input obtained from the independent astrometric calibration of different WFI frames. The solid con- tour corresponds to the half-maximum of the distribution, encompassing 62% of the points. frame. Figure 6 shows the distribution of the differences in the coordinates of 700,000 pairs of stars in patch D from independent astrometric solutions ob- sult shows that the position of the tar- FLAMES. However, even though a sat- tained for different WFI V-band frames. gets extracted from the survey data can isfactory astrometric solution was de- The rms of the distribution is ~ 0.04 arc- be directly used to position the slits and termined for the stellar field shown in sec, an estimate of the internal accura- fibers of VLT spectrographs such as Figure 5, problems are still found in cal- cy of the astrometric solution. This re- FORS1, FORS2, VIMOS, NIRMOS and ibrating crowded fields. Procedures to astrometrically calibrate these fields, including proper modelling of the optical Table 8: WFI Pilot Survey: Stellar Fields. distortions, are still being investigated. These issues as well as a full descrip- α δ Object Name Object Type 1950 1950 Filters tion of the observations, data reduction and derived products will be presented Baade Window (1) Bulge 17 55 28 –28 45 20 B, V, I, z in forthcoming papers. Baade Window (2) Bulge 17 56 04 –29 13 14 B, V, I, z Baade Window (3) Bulge 18 00 00 –29 52 13 B, V, I 5. Data Handling Baade Window (4) Bulge 18 05 33 –31 27 04 B, V, I Baade Window (5) Bulge 18 07 02 –31 46 27 V, I One of the most demanding tasks of Baade Window (6) Bulge 18 15 10 –33 58 38 V, I the EIS project has been the development of a data reduction pipeline in parallel to the observations which have NGC 6218 GC 16 47 14 –01 56 52 B, V, I used four different instruments (EMMI, NGC 6397 GC 17 40 41 –53 40 25 B, V, I SUSI2, SOFI and [email protected]) in a pe- NGC 6544 GC 18 07 20 –24 59 51 B. V, I riod of two years. Adaptation of the soft- NGC 6809 GC 19 39 59 –30 57 44 B, V, I ware to this rapidly changing environ- NGC 7089 GC 21 33 29 –00 49 23 B, V, I ment has been by far the most chal- NGC 6752 GC 19 10 51 –59 58 55 B, V, I lenging aspect of the project, which will NGC 6541 GC 18 08 02 –43 42 20 B, V, I hopefully subside in the next couple of NGC 6681 GC 18 43 12 –32 17 31 B, V, I years. This not only impacts the devel- NGC 6656 GC 18 36 24 –23 54 12 B, V, I opment effort but also the data reduc- NGC 6717 GC 18 55 06 –22 42 03 B, V, I tion and verification which depend on the stability of the pipeline. NGC 6723 GC 18 59 33 –36 37 54 B, V, I The original EIS pipeline was developed over a period of one year, starting IC 4651 OC 17 24 42 –49 57 00 B, V, I in April 1997, by assembling existing NGC 6405 OC 17 36 48 –32 11 00 B, V, I software (IRAF, SExtractor, LDAC and NGC 6208 OC 16 45 30 –53 4400 B, V, I Drizzle) adapted to the specific needs NGC 5822 OC 15 01 30 –54 09 00 B, V, I (detector and observing strategy) of the NGC 6705 OC 18 51 06 –06 16 00 B, V, I EIS-WIDE survey. The task was facilitated because most tools were devel-

43 oped by astronomers in the Member photometric zero-point, of final object 6. Data Access States who participated in this develop- catalogues, extracted from the co- ment. At the time, the data volume was added images using the most recent Unlike radio surveys and other public relatively small, which greatly facilitated and extensively tested version of surveys carried out in space, ground- the handling and processing of the SExtractor, and of colour catalogues based optical imaging surveys require data. Most of the work focused on the constructed either by the association of considerably more information to char- interfacing of the different processing objects detected in each passband or acterise the data available. An arduous blocks and in producing scripts to con- by the use of a reference image (e.g., and time-consuming task, given the trol the processing essentially in batch χ2-image). time pressure and limited personnel, mode. A lot of effort was also spent in 3. Quality control – this module car- has been to establish adequate chan- the design and implementation of a ries out the quality control of the data. It nels of communication between the database which was primarily used for monitors image distortions, computes Survey Team and the community at bookkeeping. the rms distribution of the residuals of large to describe the progress of the After the first EIS data release in the astrometric solution, compares the project and to provide the necessary in- March 1998, an upgrade was carried astrometry of the objects extracted in formation about the observations and out to incorporate ECLIPSE (Devillard, different passbands and checks the the various data products available. In Jung & Cuby 1999) for the reduction of spatial uniformity of the photometric this first phase this was accomplished SOFI data and to handle the two-CCD zero-point. It also performs a prelimi- by posting as much information as pos- mosaic of SUSI2. Even though time- nary scientific evaluation of the data. It sible on the web, by periodically re- consuming, implementing these add- culls object catalogues, computes sim- viewing the status of the project in The ons did not require major changes in ple statistics such as number counts, Messenger (da Costa 1997, da Costa the design of the pipeline. colour-colour diagrams, colour distribu- et al. 1998, da Costa & Renzini 1998) The advent of the WFI, on the other tions and the two-point angular correla- and by submitting and circulating pa- hand, required a major overhaul driven tion function for objects classified as pers with each data release. by the sheer volume of the data. New galaxies. The results are compared During the first phase, requests for hardware was required as well as a ma- with other available data to check the large volumes of data from different jor re-design of the pipeline architec- photometric zero-points, the star- groups have been processed by the ESO ture. Over the past year new machines galaxy classification and uniformity of Archive Group and about 100 Gby of data have been added, the number of the object catalogues. have been delivered in various forms. The processors available in each server in- 4. Target Selection – this module is requests originated from 11 countries, increased, disk and memory capacities used to produce a variety of target cluding all of the ESO member states. Ac- expanded and DLT tape units installed. lists. It also produces image postage cording to current guidelines, catalogues In parallel, during the past six months stamps (typically 2 × 2 arcmin) in differ- are available world-wide, but images are significant modifications have been ent passbands of the selected objects restricted to ESO member states, with made to the pipeline software, even for visual inspection. In time, other fa- the exception of those of the HDF-S though the basic algorithms remain es- cilities are expected to be added, in par- fields. Table 9 shows a summary of data sentially the same. A more sophisticat- ticular for searches for variable and requests listing the product and the ed control system has been developed proper-motion objects. Among many number of requests up to September to administer the resources of the sys- possible applications, a major goal is to 1999. The access to the EIS home page tem, to retrieve/store images in the tape define secondary-standards in the sur- has increased by a factor of 5 over library, to supervise the reduction proc- roundings of the traditional Landolt the period of the survey. It currently ess and to communicate with an ex- fields, essential to monitor the zero- averages about 30,000 hits/month, panded implementation of the database point of the different chips in the CCD- reaching peaks twice as high after a in a variety of ways. The new pipeline mosaic. data release. About one-third of these allows the reduction of both single CCD Even though the backbone of the hits are from research institutes in Eu- (SOFI and other single-chip instru- pipeline is in place, a lot of work re- rope and abroad, constantly monitor- ments) and mosaics (e.g., SUSI2, WFI) mains to be done, especially on the ing the progress of the survey, while the and parts of the code have been opti- control system, in order to allow the rest are from the general public. In this mised and parallelised to increase the process to be fully automated. Time respect, it is worth noting the popularity overall throughput to match the expect- benchmarks of different parts of the of the EIS image gallery, with about ed input data rate. Finally, the pipeline pipeline demonstrate that most of the 35,000 images being retrieved from the has been made more user-friendly and modules are operating at a rate ~ 0.2 web. A total of about 4500 data requests uniform in order to facilitate its mainte- Mpixels/sec. For instance, it currently have been processed so far, about half nance and integrity, to prevent disconti- takes about 7 hours to process 20 WFI of which asked for derived products such nuities caused by the constant changes frames and produce the results re- as target lists and catalogues. By ex- of Survey Team members and to allow quired for a quick-look release of a panding the scope of the observations to it to be run by other interested users. square degree worth of data. This cor- other areas of application and by im- Broadly speaking the pipeline con- responds to 2 hours of observations in sists of four logical blocks: the wide-angle-survey mode adopted 1. Image Processing – this module by the Pilot Survey. Even though pre- Table 9: Summary of Data Requests (Sep- performs the usual image processing liminary, these numbers show that the tember 1999). steps removing instrument signatures current pipeline is capable of coping and astrometrically and photometrically with the volume of data being produced Product Number calibrating each individual WFI frame. It by the survey. Furthermore, the pipeline of Requests then performs the co-addition of dith- is not yet fully implemented and tested ered WFI frames, producing a single and there still is considerable room for Bulk Data 50 FITS image from which object cata- improving the system throughput. The Cutouts 1128 logues are extracted. These are the ba- final consolidation of the software Mini-images 319 sic products produced automatically by should last until the end of 1999, with a Target Lists 737 the pipeline which are then transferred final target date of April 1, 2000. A de- EIS-Deep catalogues 1440 to the archive for quick-look releases. tailed description of the available soft- HDF-S images 821 2. Image Mosaicing – this module ware will be presented in the “EIS-WFI TOTAL 4495 deals with the production of image mo- Pipeline Primer” currently in prepara- Gallery Images 34538 saics, which implies adjustments of the tion by the Survey Team.

44 proving the interface with the public, these vey data. The need for such databases which will hopefully provide interested numbers are likely to increase signifi- transcends EIS and several groups in Eu- users with a broader range of products cantly. rope involved in wide-field imaging sur- as well as more flexibility and thus veys are tackling the same problem (e.g. contribute to the scientific exploitation 7. Future Developments Terapix, ΩCam). These various groups of the VLT. are exchanging ideas, and a preliminary With the recent approval of a two-year implementation has recently been test- 9. Acknowledgements Large Programme to conduct a Public ed by the EIS Survey Team, in collabo- Survey (see Renzini & da Costa 1999) ration with the ESO Archive Group, us- We would like to thank A. Renzini for with the WFI and SOFI, it is now possi- ing the Objectivity database. As the work his effort in steering the imaging survey, ble to envision a more stable operation on the pipeline is consolidated, the at- the WG members for their scientific in- leading to the production of more stan- tention of the software developers in the put and former EIS team members for dard data sets. In this framework, new Survey Team will focus on this critical their contribution over the years. Special ways of making the data accessible to the area which is a key element for the effi- thanks to E. Bertin for addressing the in- community will be explored. Quick-look cient scientific exploitation of the survey numerable questions and requests posed releases of the data, distribution of com- data. by the EIS team and for introducing new pressed data and the implementation of features to SExtractor as needed and to new on-line features are being consid- 8. Conclusions T. Erben for his help in the analysis of the ered as described by Renzini & da Cos- PSF. ta (1999). These, together with the oth- After two years of observations, soft- er products already available should pro- ware development, data reduction and References vide users with a large range of options distribution, nearly all the data from the and greater flexibility to explore the avail- ongoing Public Surveys (EIS-WIDE, Benoist et al. 1999, A&A, 346, 58. able data. EIS-DEEP and Pilot Survey) are publicly da Costa, L., 1997, The Messenger, 88, 34. da Costa. L. et al., 1998, The Messenger, 91, A considerable effort is also underway available. The only exceptions are the 49. to implement a flexible design to store the data gathered from observations con- da Costa, L. & Renzini, A., 1998, The Mes- objects extracted from the public survey ducted with the WFI in September 1999 senger, 92, 40. images in a database which will allow and some data taken with the narrow- da Costa, L. et al., l999b, A&A, in press (as- users to directly select and display objects band z filter for which no proper calibra- tro-ph/9812105). according to their positions in the sky, tion is currently available. The EIS ex- Devillard, N., Jung, Y. & Cuby, J.G., 1999, The their magnitudes and their colours with- perience has been extremely valuable for Messenger, 95, 5. out any reference to a specific catalogue. the new Public Survey to be conducted Nonino, M. et al. 1999, A&A Supp., 137, 51. The database will be a dynamic entity that in Periods 64–67, as important lessons Olsen et al. 1999a, A&A, 345, 681. Olsen et al. 1999b, A&A, 345, 363. will grow in time as information from new have been learned in handling large vol- Prandoni et al. 1999, A&A, 345, 448. observations of different sky regions umes of data, in setting up rules for au- Renzini, A. & da Costa, L., 1997, The Mes- or repeated observations of the same ob- tomatically processing images and senger, 87, 23. jects are continuously added to it. The im- recipes for data quality control, and in Renzini, A. & da Costa, L., 1999, in this issue plementation of such a database and ways and means of exporting the data of The Messenger. database tools where astronomical ob- to the users in an adequate form and in Rengelink et al., 1999, A&A, submitted (astro- jects extracted from survey data can be a timely fashion. With the accumulated ph/9812190). stored, updated and associated with all experience from four previous releases Scodeggio et al. 1999, A&A Supp., 137, 83. the information available from internal and and with the major development phase Zaggia, S. et al. 1999, A&A Supp., 137, 75. external sources is an essential element drawing to a close, new ways of making for the successful use of the Public Sur- the data available are being considered E-mail: ldacosta

45 OBSERVERS WITH THE VLT

VLT/ISAAC and HST/WFPC2 Observations of NGC 3603 B. BRANDL1, W. BRANDNER2, E.K. GREBEL3 AND H. ZINNECKER 4*

1Cornell University, Ithaca; 2University of Hawaii, Honolulu; 3University of Washington at Seattle; 4Astrophysikalisches Institut Potsdam

1. Abstract spectral signatures of Wolf-Rayet (W-R) side (Churchwell et al. 1987; O’Dell et al. stars contribute more than 2000 Mᓃ to 1993). FUV photons (13.6 eV > hν > 6 We have studied NGC 3603, the the cluster mass. Normally, W-R stars are eV) heat up the inside of the proplyd en- most massive visible HII region in the evolved supergiants that have long left velope and lead to the dissociation of mol- Galaxy, with VLT/ISAAC in the near-in- the main sequence and have ages of 3–5 ecules in the outer layers of the circum- frared (NIR) Js, H, and Ks-bands and Myr. In NGC 3603, however, the W-R stellar disk (Johnstone et al. 1998). The HST/WFPC2 at Hα and [N II] wave- stars also show hydrogen absorption resulting evaporation flow provides a lengths. In this Messenger article we lines in addition to the typical W-R fea- steady supply of neutral atoms to the ion- describe the data analysis and some tures. It is believed that these stars are isation front and leads to the development first results from both our complemen- still main-sequence, core hydrogen- of a cometary tail (McCullough et al. 1995; tary observations. burning stars that are so massive and so Störzer & Hollenbach 1999). Until re- Our HST/WFPC2 gave us an un- close to the Eddington limit that they are cently, only one other proplyd had been precedented high-resolution view of loosing their outer envelopes through fast found outside the Orion nebula. It is lo- the interstellar medium in the giant HII winds, and thus resemble evolved W-R cated in the vicinity of the O7V star Her- region and its ionisation structure. stars. In comparison to the Orion Trapez- schel 36 in the Lagoon Nebula (M8, Among the findings are two gigantic ium system, NGC 3603 with its more than Stecklum et al. 1998). gas columns (similar to the famous 50 O and W-R stars producing a Lyman “elephant trunks” in M16) and three pro- continuum flux of 1051 s–1 (Kennicutt 3. VLT/ISAAC Observations plyd-like structures. The emission neb- 1984; Drissen et al. 1995) has about 100 ulae are clearly resolved; all three neb- times the ionising power of the Trapezi- We observed NGC 3603 through the ulae are tadpole shaped, with the bright um cluster. Js = 1.16–1.32 µm, H = 1.50–1.80 µm, 7 ionisation front at the head facing the With a bolometric luminosity Lbol > 10 and Ks = 2.03–2.30 µm broadband filters central cluster and a fainter ionisation Lᓃ, NGC 3603 has about 10% of the lu- using the NIR camera ISAAC on ANTU, front around the tail pointing away from minosity of 30 Doradus and looks in many the first VLT unit telescope. The obser- the cluster. Typical sizes are 6000 A.U. respects very similar to its stellar core vations were made during the 4 nights of × 20,000 A.U. The nebulae share the R136 (Brandl et al. 1996). In fact, it has April 4–6 and 9, 1999, in service mode overall morphology of the proplyds been called a Galactic clone of R 136 but when the optical seeing was equal to, (“PROto PLanetarY DiskS”) in Orion, without the massive surrounding cluster or better than, 0.4″ in 1-minute expo- but are 20 to 30 times larger in size. halo (Moffat, Drissen & Shara 1994). In sures. Such seeing was essential for The VLT observations are the most many ways NGC 3603 and R136 can be accurate photometry in the crowded sensitive near-infrared observations regarded as representative building cluster and increased our sensitivity made to date of a dense starburst blocks of more distant and luminous star- to the faintest stars. The majority of region, allowing us to investigate with un- burst galaxies (Brandl, Brandner & Zin- our data were taken under photometric precedented quality its low-mass stellar necker 1999, and references therein). So conditions. population. Our sensitivity limit to stars the best way to study collective star for- Our observing strategy was to use detected in all three bands corresponds mation in a violent environment is by ex- the shortest possible frame times of to 0.1 Mᓃ for a pre-main sequence star amining these regions on a star-by-star 1.77 seconds to keep the number of of age 0.7 Myr. Our observations clear- basis. Until recently, observations of the saturated stars to a minimum. However, ly show that sub-solar-mass stars down entire stellar populations were limited to due to the system’s excellent sensi- to at least 0.1 Mᓃ do form in massive star- the massive stars by sensitivity or wave- tivity, about two dozen of the brightest bursts. length restrictions. The formation of low- stars ended up being saturated. Never- mass stars in starburst regions is of par- theless, this does not impose a prob- 2. Introduction ticular interest. Fundamental questions lem to our study of the low-mass stars. that arise in this context are: Does the Thirty-four short exposures were co- NGC 3603 is located in the Carina spi- slope of the IMF vary on small scales? added to an effective one-minute expo- ral arm (RA = 11h, DEC = –61°) at a dis- Do low-mass stars in a starburst event sure, the minimum time per pointing tance of 6–7 kpc. It is the only massive, form together with the most massive stars required to stabilise the telescope’s ac- Galactic HII region whose ionising cen- or do they form at different times or on tive optics control system. Between the tral cluster can be studied at optical wave- different timescales? And finally, one 1-minute pointings we moved the tele- lengths, due to only moderate (mainly might even ask if low-mass stars form at scope by up to 20″ offsets in a random foreground) extinction of AV = 4–5 mag all in such environments? pattern. This approach has several ad- (Moffat 1983; Melnick et al. 1989). The In addition, NGC 3603 is an ideal lab- vantages: OB stars (> 10 Mᓃ) and stars with the oratory to study individual (!) star forma- • Enlargement of the observed field of tion. Some young stars in the vicinity of view (FOV) with maximum signal-to-noise massive stars are surrounded by par- (S/N) in the cluster centre. *Our collaborators on the project described in this tially ionised circumstellar clouds with • Reduction of residual images and article are Y.-H. Chu, H. Dottori, F. Eisenhauer, A.F.J. Moffat, F. Palla, S. Richling, H.W. Yorke and S. D. cometary shape, so-called protoplanetary other array artefacts, using the median Points. disks (“Proplyds”), ionised from the out- filtering technique.

46 • Derivation of the “sky” from the target exposures using the median filtering technique. No additional time for “blank” sky frames outside the cluster was required. The sky frames have been computed using between 15 and 37 subsequent exposures per waveband and night, and careful eye-inspection showed that all sources have been efficiently removed using our modified median filtering technique which returns the lower 1/3 instead of the mean (1/2) value. We subtracted the sky background and flat-fielded each exposure using the twilight flat-fields provided by ESO. The relative position offsets were derived from cross-correlat- ing the images; the exposures were co- aligned on a 0.5 × 0.5-pixel sub-grid for better spatial resolution, and then added together us- Figure 1: Three-colour image of NGC 3603 composed from Js (blue), H (green), and Ks-band (red) images. Intensi- ing the median fil- ties are scaled in logarithmic units. The field of view is 3.′4 × 3.′4. North is up, East to the left. The insert to the lower tering technique. right is a blow-up of the central parsec2. The resulting images are 3.′4 × 3.′4 in size with pixels of 0.″074. The effective details). About 1′ south of the central observations and the systematic pho- exposure times of the final broadband im- cluster, we detect the brightest mem- tometric errors are dominated by un- ages in the central 2.′5 × 2.′5 are 37, 45, bers of the deeply embedded pro- certainties in the aperture offsets. (A and 48 minutes in Js, H, and Ks, re- tocluster IRS 9. detailed error analysis will be part spectively. In order to derive the photometric of a subsequent paper). Comparing Figure 1 shows the impressive 3-colour fluxes of the stars we used the IRAF our photometric fluxes of numerous composite image that has recently been implementation of DAOPHOT (Stetson sources with the fluxes derived by the subject of an ESO press release. 1987). We first ran DAOFIND to detect Eisenhauer et al. (1998) yields a m (http://www.eso.org/outreach/press-rel/pr- the individual sources, leading to systematic offset of 0.1 in Js and m 1999/pr-16-99.html). The brightest star in Ȃ 20,000 peaks in each waveband. 0.05 in Ks. the FOV (80″ northeast of the core) is the Many of these may be noise or red supergiant IRS 4 (Frogel, Persson, peaks in the nebular background and & Aaronson 1977). To the south of the appear only in one waveband. In order 4. HST/WFPC2 Observations cluster is a giant molecular cloud to reject spurious sources, we required (GMC). Ionising radiation and fast stel- that sources be detected independ- On March 5, 1999 we obtained lar winds from the starburst cluster are ently in all three wavebands, and that deep narrow-band Hα (F656N, 2 × excavating large gaseous pillars. Lo- the maximal deviation of the source 500s) and [NII] (F658N, 2 × 600s) cated about 20″ to the north of the position centroid between different observations. The Planetary Camera cluster centre is the blue supergiant wavebands be less than 0.075″. The (PC2) chip was centred on the bipo- Sher 25. This supergiant is unique be- resulting source list contains 6967 ob- lar outflow structure around the blue cause its circumstellar ring and bipo- jects in the entire FOV. We then flux- supergiant Sher 25. The three Wide lar outflows form an hourglass struc- calibrated the images using the faint Field Camera (WF) chips covered ture similar to that of SN 1987A (Brand- NIR standard stars from the lists by the central cluster and the HII region ner et al. 1997a, 1997b). The image also Hunt et al. (1998) and Persson et al. to the South of Sher 25. In addition, shows the three proplyd-like objects (1998). Because of the stringent re- we retrieved and analysed archival that have recently been discovered by quirements on the seeing, the PSF HST data, which had originally been Brandner et al. (2000) (see Fig. 2 for did not noticeably change during our obtained in July 1997 (PI Drissen). The

47 Figure 2: WFPC2 observations of NGC 3603. North is up and East is to the left. The upper part of the image consists of the archive data with the following colour coding: F547M (blue), F675W (green), F814W (red). Overlaid are our new WFPC2 data with the F656N data in the red channel, the average of F656N and F658N in the green channel, and F658N in the blue channel. The location of the three proplyd-like emission nebulae is indicated. The insert at the lower right is a combination of WFPC2 F656N (blue) and F658N (green) and VLT/ISAAC Ks (red) observations.

PC2 was centred on the cluster, and the F656N data in the red channel, 5. Results and Interpretation the three WF chips covered the area the average of F656N and F658N in north-west of the cluster. We combined the green channel, and F658N in the All three proplyds are tadpole shaped individual short exposures in F547M blue channel. The locations of the and rim brightened, with the extended (8 × 30s), F675W (8 × 20s), and F814W proplyds are marked by small boxes, tails facing away from the starburst (8 × 20s) to produce images with effec- and enlargements of the boxes are cluster. The portion of the ionised rims tive exposure times of 240s, 160s, and shown in the upper part of Figure 2. pointing towards the cluster are 160s, respectively. Proplyd 3 has only been observed in brighter than the rims on the opposite The HST/WFPC2 observations are intermediate and broad-band filters side. The central parts of the proplyds presented in Figure 2. The figure and thus stands out less clearly against are fainter than the rims, with a notice- shows an overlay of two composite the underlying background when com- able drop in surface brightness be- colour images. The upper part of the pared to Proplyds 1 and 2. The insert tween the head and the tail. Proplyds 2 image consists of the archive data with at the lower right shows a colour com- and 3 exhibit a largely axisymmetric the following colour coding: F547M posite of HST/WFPC2 F656N (blue) morphology, whereas Proplyd 1, which (blue), F675W (green), F814W (red). and F658N (green) data and VLT/ is also the one closest to the clus- Overlaid are our new WFPC2 data with ISAAC Ks data (red). ter, has a more complex structure. Un-

48 like the convex shape of the heads of the rived from our VLT data. The left plot con- the investigated mass range by about other proplyds, Proplyd 1 has a heart- tains all stars detected in all 3 wavebands one order of magnitude toward smaller shaped head with a collimated, outflow- within the entire FOV of 3.′4 × 3.′4 (6 pc masses and covers also a much larger like structure in between. One possible × 6 pc). Since NGC 3603 is located in the area. explanation for the more complex mor- Galactic Plane we expect a significant Because of Ȃ 10 magnitudes range phology of Proplyd 1 might be that it is contamination from field stars. To reduce in luminosities in the crowded core re- m actually a superposition of two (or maybe this contribution we followed a statistical gion, our sensitivity limits of Js Ȃ 21 , even three) individual proplyds or that approach by subtracting the average H Ȃ 20m and Ks Ȃ 19m don’t appear the photoevaporative flows of several number of field stars found in the regions to be exceedingly faint (and are about disks in a multiple system interact to around the cluster at r > 75″ (central plot) 3 magnitudes above ISAAC’s detection produce this complex single structure. per magnitude and per colour bin (0.5 limit for isolated sources). However, only At distances of 7.4″ and 2.9″ from Pro- mag each). VLT/ISAAC’s high angular resolution, plyd 1 and 2, respectively, faint arc-like The accuracy of our statistical sub- PSF stability, and overall sensitivity en- Hα emission features are seen on the traction is mainly limited by three factors: abled us to study the sub-solar stellar WFPC2 frames. The arcs are located in first, we cannot rule out that low-mass population in a starburst region on a star- the direction of the cluster, and may be pre-main-sequence stars are also pres- by-star basis. the signatures of bow shocks created by ent in the outskirts of the cluster. Sec- the interaction of proplyd winds with the ond, because of crowding on one hand Acknowledgements: We would like winds from the massive stars in the cen- and dithering which leads to shorter to thank the ESO staff, in particular those tral cluster. Additional faint filaments lo- effective integration times outside the involved in the service observations, for cated between the nebulae and the cen- central 2.′5 × 2.′5 on the other hand, the their excellent work. tral ionising cluster can be interpreted as completeness limit varies across the bow shocks resulting from the interaction FOV. Third, local nebulosities may hide of the fast winds from the high-mass stars background field stars. However, none References in the cluster with the evaporation flow of these potential errors affects our from the proplyds. conclusions drawn from the CMD. The Brandl, B., et al. 1996, ApJ, 466, 254. Low-resolution spectra obtained with resulting net CMD for cluster stars with- Brandl, B., Brandner, W. & Zinnecker, H. 1999, EFOSC2 at the ESO/MPI 2.2-m tele- in r < 33″ of NGC 3603 is shown at the Star Formation 1999, eds. T. Nakamoto et scope of the brightest nebula, which is right in the figure on page 1. We over- al., in press. at a projected separation of 1.3 pc from laid the theoretical isochrones of pre- Brandl, B., et al. 1999, A&A Letters, 352, L69. the cluster, reveal that it has the spec- main sequence stars from Palla & Brandner, W., et al. 1997a, ApJ, 475, L45. tral excitation characteristics of an ul- Stahler (1999) down to 0.1 Mᓃ. We as- Brandner, W., et al. 1997b, ApJ, 489, tra compact H II region with electron sumed a distance modulus of (m–M)ᓃ L153. densities well in excess of 104 cm–3. = 13.9 based on the distance of 6 kpc Brandner, W., et al. 2000, AJ, in press (Jan- The near-infrared data reveal a point (De Pree et al. 1999) and an average uary 2000 issue). m source superimposed on the ionisation foreground extinction of AV = 4.5 fol- Churchwell, E., Felliw, M., Wood, D.O.S. & front. lowing the reddening law by Rieke & Massi, M. 1987, ApJ, 321, 516. The striking similarity of the tadpole- Lebofski (1985). De Pree, C.G., Nysewander, M.C. & Goss, shaped emission nebulae in NGC 3603 Applying the isochrones to measured W.M. 1999, AJ, 117, 2902. Dottori, H., et al. 2000, subm. to A&A. to the proplyds in Orion suggests that magnitudes could be misleading since Drissen, L., Moffat, A.F.J., Walborn, N. & the physical structure of both types of ob- the theoretical calculations include only Shara, M.M. 1995, AJ, 110, 2235. jects might be the same. Our hydro- stellar photospheres while the stars still Eisenhauer, F., Quirrenbach, A., Zinnecker, H. dynamical simulations (Brandner et al. may be surrounded by dust envelopes & Genzel, R. 1998, ApJ, 498, 278. 2000) reproduce the overall morpholo- and accretion disks. This would lead to Frogel, J.A., Persson, S.E., Aaronson, M. gy of the proplyds in NGC 3603 very well, excess emission in the NIR and make 1977, ApJ, 213, 723. but also indicate that mass-loss rates of the stars appear younger than they ac- Hunt, L.K., et al. 1998, AJ, 115, 2594. up to 10–5 Mo yr–1 are required in order tually are. Typical excess emission of Johnstone, D., Hollenbach, D. & Bally, J. 1998, to explain the size of the proplyds. Due classical T-Tauri stars in the Taurus- ApJ, 499, 758. Kennicutt, R.C. Jr. 1984, ApJ, 287, 116. to these high mass-loss rates, the pro- Auriga complex have been determined m m McCullough, P.R., Fugate, R.Q., Christou, J.C. plyds in NGC 3603 should only survive as DH = 0.2 , and DK = 0.5 (Meyer, Et al. 1995, ApJ, 438, 394. 5 10 yrs. Despite this short survival time, Calvet & Hillenbrand 1997). The upper Melnick, J., Tapia, M. & Terlevich, . 1989, A&A, we detect three proplyds. This indicates part of the cluster-minus-field CMD 213, 89. that circumstellar disks must be common clearly shows a main sequence with a Meyer, M.R., Calvet, N., & Hillenbrand, L.A. around young stars in NGC 3603 and that marked knee indicating the transition to 1997, AJ, 114, 288. these particular proplyds have only repre-main-sequence stars. The turn-on Moffat, A.F.J. 1983, A&A, 124, 273. cently been exposed to their present occurs at J Ȃ 15.5 mag (m Ȃ 2.9 Moffat, A.F.J., Drissen, L., & Shara, M.M. 1994, s ApJ, 436, 183. harsh UV environment. Mᓃ). Below the turn-on the main-se- O'Dell, C.R.O., Wen, Z. & Hu, X. 1993, ApJ, The point source close to the head of quence basically disappears. We note 410, 686. Proplyd 3 is already detected on the that the width of the pre-main se- Palla, F. & Stahler, S. W. 1993, ApJ, 418, broadband HST/WFPC2 observations quence in the right part of the figure on 414. (see Figure 2). The WFPC2 images show page 1 does not significantly broaden Palla, F. & Stahler, S.W. 1999, ApJ, in press that the point source is actually located toward fainter magnitudes, indicating (20 Nov). in front of Proplyd 3 and thus very likely that our photometry is not limited by Persson, S.E. et al. 1998, AJ, 116, 247. not physically associated with it. The in- photometric errors. The scatter may in Stecklum, B., Henning, T., Feldt, M., et al. frared source in Proplyd 1 is also detected fact be real and due to varying fore- 1998, ApJ, 115, 767. as an underlying, heavily reddened con- ground extinction, infrared excess and Stetson, P.B. 1987, PASP, 99, 191. Störzer, H. & Hollenbach, D. 1999, ApJ, 515, tinuum source in the spectrum. A more evolutionary stage. In that case the left 669. detailed analysis of ground based opti- rim of the distribution would be repre- Zinnecker, McCaughrean & Wilking 1993, cal images and spectra of the proplyds sentative of the “true” colour of the Protostars and Planets III, eds. E.H. Levy and other compact nebulae in NGC 3603 most evolved stars. Fitting isochrones & J.I. Lunine (University of Arizona Press), will be presented in Dottori et al. (2000). to the left rim in the CMD yields an 429. The figure on page 1 of this issue of age of only 0.3–1.0 Myr. Our result is The Messenger shows the resulting in good agreement with the study by colour-magnitude diagrams (CMD) de- Eisenhauer et al. (1998) but extends E-mail: [email protected]

49 From Intracluster Planetary Nebulae to High- Redshift Lyα Emitters R.P. KUDRITZKI and R.H. MÉNDEZ, Munich University Observatory J.J. FELDMEIER and R. CIARDULLO, Dept. of Astron. and Astrophys., Penn State University G.H. JACOBY, Kitt Peak National Observatory, Tucson K.C. FREEMAN, Mt. Stromlo and Siding Spring Observatories M. ARNABOLDI and M. CAPACCIOLI, Osservatorio Astronomico di Capodimonte, Napoli O. GERHARD, Astronomisches Institut, Universität Basel H.C. FORD, Physics and Astronomy Dept., Johns Hopkins University, Baltimore

1. Introduction and only by chance they appear pro- essary for reliable abundance determi- jected upon NGC 4406. Direct evi- nations, at least for a few of the bright- The fact that the spectra of planetary dence of the existence of red giants est intracluster PN candidates. This nebulae (PNs) are dominated by strong belonging to this Virgo intracluster stel- was not to be; in what follows we report emission lines, with a very weak contin- lar population was subsequently re- the unexpected outcome of our first uum, has two interesting consequences ported by Ferguson et al. (1998), who VLT run. Here we will restrict our de- for extragalactic work. First, PNs in oth- worked with the Hubble Space Tel- scription to the LPF, where the most in- er galaxies are easily detectable: we escope. These red giants are too faint teresting objects are located. A paper need two images of a galaxy, one taken for spectroscopic studies. In contrast, with a complete report has been sub- through a narrow-band filter transmit- the intracluster PNs offer an ideal op- mitted to ApJ. ting a strong nebular emission (e.g. portunity to learn about the kinemat- [O III] λ5007) and another taken ic behaviour and perhaps even the 3. Observations and First through an off-band filter which does abundances of the diffuse stellar popu- Results not transmit any strong nebular line. In lation. A search for intracluster PNs in the on-band image the PN appears as different positions across the Virgo In April 1999 we used the VLT UT1 a point source. Since the nebular con- cluster started immediately and pro- with FORS in multi-object spectroscop- tinuum is so weak, the point source duced dozens of PN candidates ic mode (MOS) with grism 300V. We should be invisible in the off-band im- (Méndez et al. 1997, Feldmeier et al. defined 1 arcsec wide slitlets for sever- age. By blinking the two images, and 1998). al PN candidates in the LPF, and added provided that the off-band image is The “on-band/off-band” narrow-band slitlets for 2 objects suspected to be deep enough, it is easy to identify the filter technique used to discover the PN QSOs or starbursts because they were PNs. Second, for each detected PN we candidates allows the detection of a visible (although much weaker) in the can measure an accurate radial veloci- single emission line. This is not neces- off-band discovery image (Méndez et ty, because all the flux we have detect- sarily the desired [O III] λ5007; it might al. 1997). Some other slitlets were ed is concentrated at the redshifted be another emission line at higher z, placed on stars or galaxies in the field, wavelength of the emission line. Thus redshifted into the on-band filter, like in order to check the slitlet positioning we can take a good spectrogram in an [O II] λ3727 at z = 0.35, or Lyα at z = (done by taking a short exposure with- exposure time not much longer than 3.14. In previous work (e.g. Méndez et out grism through the slitlets) and to what we need for the imaging. Since al. 1997) we argued that most of our help locate the dispersion lines as a PNs are preferentially discovered in detections had to be real PNs because function of position across the field, the outskirts of galaxies, where the (a) the surface density of emission-line which is important for the correct ex- surface brightness is lower, this means galaxies derived from previous studies traction of spectra consisting of isolated that PNs are ideal test particles for (for example Thompson et al. 1995) emission lines. studies of rotation and mass distri- was not high enough to explain all the After completing 5 exposures of the bution in the halos of the corresponding detections; (b) the luminosity function initial MOS configuration, each 40 min- galaxies. Typical examples of such of the detected sources, if we assume utes long, the surprising result was that studies are Hui et al. (1995) on NGC them to be at the distance of the Virgo our targets in the LPF were confirmed 5128, and Arnaboldi et al. (1996) on the cluster, is in good agreement with the as emission-line objects, but none of Virgo cluster galaxy NGC 4406. Here PN luminosity functions derived in them is a PN. If we expose long begins an unusual chain of unexpected several Virgo galaxies (Jacoby et al. enough, a PN must show not only [O III] results. 1990). λ5007, but also [O III] λ4959, which is Then we decided to take deep three times weaker. In none of our 2. Intracluster PNs VLT+FORS spectrograms of several in- sources could we find the second [OIII] tracluster PN candidates, located in line. The spectra show only a strong, The giant galaxy NGC 4406 is char- Field 1 of Feldmeier et al. (1998) and in isolated and narrow emission line at acterised by a negative redshift (–230 the smaller field observed by Méndez et wavelengths from 5007 to 5042 Å (en- km s–1), and indeed many of its PNs, al. (1997; in what follows the “La Palma compassing the full width of the on- discovered by Jacoby et al. (1990) and Field” or LPF, because the discovery band discovery filter), and nothing measured by Arnaboldi et al. (1996), images were taken with the 4.2-m else. have the expected redshifts. However, William Herschel Telescope at La Of the 2 QSO or starburst candi- 3 of these PNs turn out to have red- Palma). Our purpose was first of all to dates, one was confirmed as a QSO at shifts typical of the Virgo cluster measure radial velocities and provide z = 3.13, showing a typical broad-lined (around 1300 km s–1). The conclusion spectroscopic confirmation of the PN Lyα and several other emissions, in- was that these 3 PNs do not belong to nature of the candidates, since a few of cluding C IV λ1550. The other one ap- NGC 4406; they must be members of them could be high-redshift sources. pears to be a starburst region (one an intergalactic or intracluster diffuse Second, we were hoping to detect the strong, isolated and narrow emission stellar population in the Virgo cluster, very faint diagnostic emission lines nec- line, with faint continuum).

50 In this way we found ourselves un- able to work on the expected intracluster PNs, and facing a different problem, namely, we had to identify the isolated emission. Was it [O II] λ3727 at z = 0.35, or could it be Lyα at z = 3.13? We defined a new MOS configuration, placing the slitlets so as to cover as much wavelength range towards the infrared as possible. We were able to take three exposures with this new MOS configuration. Based on the absence of any other emission line in our spectra, we could rule out all the alternatives, e.g. if the detected emission were [O II] λ3727 redshifted into the on-band filter, then we should see other features, like Hβ, the [O III] lines and Hα, at the corresponding redshift. A similar argument can be Figure 1. A section of an image produced by coadding 3 MOS exposures of the La Palma Field used to reject Mg II λ2798 at z = 0.79. (LPF), taken with FORS. Here 4 of the 19 FORS slitlets are shown. These spectra, taken during new moon, are dominated by strong sky emissions. From top to bottom: (1) one of our in- There is only one possible identifica- α tion: Lyα at z = 3.13. We had unwilling- tracluster PN candidates, now re-identified as a high-redshift Ly emitter; (2) a foreground star; (3) a low-redshift galaxy showing a bright H II region at the edge, where Hβ and the [OIII] emis- ly found a nice collection of high-red- λλ α sions 4959, 5007 are visible; (4) one of the QSO or starburst candidates, confirmed as a shift Ly emitters, which was an ade- QSO. The strong and broad emission line is Lyα at z = 3.13. A weak continuum is visible. C quate compensation for the loss of the IV λ1550 is visible to the right, flanked by strong sky emissions. expected intracluster PNs.

4. How Could This Happen? equivalent width, between 20 and 200 correspond to the formation of rather Å in the rest frame. For these typical small galaxies, some of which are Not just a few, but all the sources we equivalent widths the most probable perhaps destined to merge. However, tested with FORS are high-redshift explanation is H II regions ionised by one of our sources might become a galaxies. However it would be wrong to recently formed massive stars (e.g. serious candidate for a proto-giant conclude that intracluster PNs do not Chariot and Fall 1993). The observed spheroidal galaxy if we assumed con- exist. In fact, multi-object spectroscopy Lyα fluxes are between 2 × 10–17 and 2 tinuous star formation, a low mass cut- with the AAT and the 2-degree-field × 10–16 erg cm–2 s–1. Adopting z = 3.13, off of 0.1 Mᓃ in the IMF, and a flat ac- –1 –1 Ω (2df) fibre spectrograph has confirmed H0 = 70 km s Mpc and q0 = 0.5 we celerating universe with 0 = 0.2 and the PN nature of many Virgo intraclus- get a luminosity distance of 1.8 × 104 ΩΛ = O.8. ter candidates, by detecting both [O III] Mpc, which implies Lyα luminosities for The implied star formation density emissions at 4959 and 5007 Å (Free- our sources between 2 × 108 and 2 × in our sampled comoving volume is 9 man et al. 1999). What happened to 10 Lᓃ. These numbers depend on our probably somewhat smaller than, but us is that the surface density of intra- assumptions about the cosmological of the same order of magnitude as, cluster PNs in the LPF appears to be parameters: for a flat universe with the star formation density at z ~ 3 de- much lower than in other places across Ω0 = 0.2, ΩΛ = 0.8, and the same z rived by other authors from Lyman- the Virgo cluster, probably indicating and H0 as above, the luminosity dis- break galaxy surveys (see e.g. Steidel some degree of clumpiness in the distance becomes 3 × 104 Mpc, and the et al. 1999). This result agrees with the tribution of the diffuse intracluster pop- Lyα luminosities become larger by a expectation that the Lyα emitters are ulation. This will be verified by wide field factor 2.8. Our Lyα emitters are quite a low-metallicity (or low-dust) tail in a surveys currently in progress. similar to those recently found by Hu distribution of star-forming regions at There is another aspect to consider: (1998), Cowie & Hu (1998) and Hu et high redshifts. Finally, the Lyα emitters the rather small area of the LPF (50 al. (1998) at redshifts from 3 to 5. may contribute as many H-ionising pho- arcmin2). Given a small area and a low Using simple population synthesis tons as QSOs and Ly-break galaxies at surface density, the total number of in- models, on the assumption that these z ~ 3 (see e.g. Madau et al. 1998). They tracluster PNs is small, and it becomes sources are produced by star for- are therefore potentially significant for very improbable to find bright PNs. It mation with a Salpeter initial mass the ionisation budget of the early uni- seems that the fraction of high-redshift function and a lower mass cut-off of verse. Lyα sources in the on-band/off-band 0.5 Mᓃ, we have concluded (Kudritzki samples is higher at faint magnitudes, et al. 1999 ApJ submitted) that the 6. Future Work thus explaining the detections in the nebulae are nearly optically thick and LPF. We conclude that it is more effi- must have a very low dust content, in Now we would like to do some cient to search for intracluster PNs us- order to explain the high observed additional work on these Lyα emitters. ing wide-field survey cameras, like the Lyα equivalent widths. These galaxies In particular we would like to use WFI at the ESO 2.2-m telescope. With must differ substantially from the typ- ISAAC; we expect the strongest [O II] a larger area the total number of PNs ical Lyman break galaxies detected and [O III] forbidden lines as well as increases and it is more probable to find through their U-dropout; those are Hβ to be detectable with the Short bright ones; at brighter magnitudes the characterised by a strong continuum Wavelength arm of ISAAC in Medium fraction of high-redshift Lyα emitters is (needed for their very detection) and Resolution spectroscopic mode. This smaller. weak or absent Lyα emission. We at- would allow us to either determine tribute this difference to a lower dust or (in the case of non-detection) put 5. Properties of the Lyα Emitters content of our objects. an upper limit to the metallicity. It might For the cosmological and star forma- also be possible to clarify to what The weak (in most cases undetect- tion parameters we adopted, the total extent the production of Lyα photons ed) continuum implies a large Lyα stellar mass produced would seem to can be attributed to AGN activity in-

51 stead of recent massive star formation. [4] Feldmeier J.J., Ciardullo R., Jacoby G.H. [10] Jacoby G.H., Ciardullo R., Ford H.C. The firm detection of an infrared contin- 1998, ApJ 503, 109. 1990, ApJ 356, 332. uum would help to constrain the char- [5] Ferguson H.C., Tanvir N.R., von Hippel [11] Madau P., Haardt F., Rees M.J. 1998, acteristics and total mass of the stellar T. 1998, Nature 391, 461. preprint astro-ph/9809058 (4 Sep 1998). populations through comparison with [6] Freeman K.C. et al., 1999, in Galaxy [12] Méndez R.H., Guerrero M.A., Freeman Dynamics: from the Early Universe to K.C. et al. 1997, ApJ 491, L23. population synthesis models. the Present, ASP Conf. Ser., eds. F. [13] Steidel C.C., Adelberger K.L., Giavalis- Combes, G.A. Mamon, V. Charmandaris, co M. et al. 1999, preprint astroph/ References in press, astro-ph/9910057. 9811399v2 (23 Jan 1999). [7] Hu E.M. 1998, ASP Conf Ser. 146, p. 148. [14] Thompson D., Djorgovski S., Trauger J. [1] Arnaboldi M., Freeman K.C., Méndez [8] Hu E.M., Cowie L.L., McMahon R.G. 1995, AJ 110, 96. R.H. et al. 1996, ApJ 472, 145. 1998, ApJ 502, L99. [2] Charlot S., Fall S.M. 1993, ApJ 415, 580. [9] Hui X., Ford H.C., Freeman K.C., Dopi- [3] Cowie L.L., Hu E.M. 1998, AJ 115,1319. ta M.A. 1995, ApJ 449, 592. E-mail: [email protected]

VLT Spectroscopy of the z = 4.11 Radio Galaxy TN J1338–1942

C. DE BREUCK1,2, W. VAN BREUGEL2, D. MINNITI 2,3, G. MILEY1, H. RÖTTGERING1, S.A. STANFORD2 and C. CARILLI 4

1Sterrewacht Leiden, The Netherlands (debreuck,miley,[email protected]) 2Institute of Geophysics and Planetary Physics, Lawrence Livermore National Laboratory, Livermore, U.S.A. (wil,[email protected]) 3P. Universidad Católica, Santiago, Chile ([email protected]) 4National Radio Astronomy Observatory, Socorro, USA ([email protected])

High-redshift radio galaxies (HzRGs) most massive galaxies at high red- stars were formed. The near-IR Hub- play an important role in cosmology. They shifts, and can therefore constrain the ble K – z diagram of powerful radio are likely to be some of the oldest and epoch at which the first generation of galaxies shows a remarkably tight correlation from the present time out to z = 5.19, despite significant K-corrections (van Breugel et al. 1998, 1999). This indicates that we can follow the evolution of the hosts of HzRGs from near their formation epoch out to low redshift (z <~ 1), where powerful radio sources inhabit massive elliptical galaxies (e.g. Lilly & Longair 1984; Best, Longair & Röttgering 1998). For example, at z ~ 3, we observe a change in the observed K-band morphologies of HzRGs from large-scale low-surface brightness emission with bright radio-aligned clumps at z <~ 3 to smooth, compact structures, sometimes showing elliptical shapes, like their ocal Universe counterparts (van Breugel et al. 1998). These surrounding clumps have properties similar to the UV dropout galaxies at similar redshifts (Pen- tericci et al. 1999), and indicate that HzRGs often reside in (proto-)cluster environments. This evolution picture seems consistent with the hierarchical clustering formation models (e.g. Kauff- mann et al. 1999), where massive objects form by accretion of smaller systems located in over-dense regions.

Figure 1: VLT spectrum of TN J1338–1942. The lower panel has been boxcar smoothed by a factor of 9 to better show the shape of the Lyα forest and the Lyman limit.

52 Detailed studies of HzRG can then be compared with the predictions of the hierarchical clustering scenarios. Using newly available radio surveys, we have begun a systematic search for z > 3 radio galaxies (De Breuck et al. 2000), to be followed by more detailed studies of selected objects. During the second week of VLT Antu operations in April 1999, we used FORS1 to obtain a spectrum of the second highest redshift radio galaxy from our sample, TN J1338–1942 at z = 4.11. We discovered this object in March 1997 using the ESO 3.6-m telescope (De Breuck et al. 1999a). The purpose of these VLT observations was to study the Lyα emission and the UV- continuum. We used the 600R grism with a 1.3″ wide slit, and integrated for 47 min. The spectrum, shown in Figures 1 and 2, is dominated by the bright Lyα line (FLyα = 2 × 10–15 erg s–1 cm–2) with a rest-frame equivalent width of 210±50Å, a typical value for HzRGs. Most notable from Figure 2 is the large asymmetry in the emission profile, consistent with a very wide (~ 1400 km/s) blueward depression over the entire spatial extent of the emission (~10 kpc). Similar asymmetries have been detected in other HzRGs, and have been interpreted as due to absorption by cold HI gas in a halo sur- Figure 2: Part of the spectrum around the Lyα line. The solid line is the model consisting of a rounding the radio galaxy (e.g. van Ojik Gaussian emission profile (dashed line) and a Voigt absorption profile with the indicated pa- et al. 1997; Dey 1999). The Lyα profile rameters. is well fit by a simple model consisting of a Gaussian emission profile and a single Voigt absorption function (see ments, our results suggest that the fore- Bibliography Figure 2). This model constrains the ground HI absorption (≡ DA) in quasars HI column density in the range 3.5 × could have been overestimated be- [1] Best, P. N., Longair, M. N., & Röttgering, 1019 – 1.3 × 1020 cm–2. From this, we cause the highest redshift (z > 4) qua- H.J.A. 1998, MNRAS, 295, 549. derive the total mass of the absorber sars have been found from samples [2] De Breuck, C., van Breugel, W., Röttger- 7 ing, H., Miley, G., & Carilli, C. 1999a, in at 2 – 10 × 10 Mᓃ, which is compara- with an optical colour selection, thereby ble or slightly less than the total HII excluding quasars with relatively small “Looking Deep in the Southern Sky”, ed. α breaks across Lyα. R. Morganti & W. Couch (Berlin Heidel- mass, as derived from the Ly emis- berg: Springer), p. 246. sion (De Breuck et al. 1999b). These It is therefore particularly important [3] De Breuck, C. van Breugel, W., Minniti, results show that TN J1338–1942 is to expand the number of objects at D., Miley, G., Röttgering, H., Stanford, S. surrounded by a large reservoir of z > 4 selected by techniques other A., & Carilli, C. 1999 A&A, 1999b, 352, gas. than optical colours. Radio galaxies L51. We also detect continuum emission are the brightest objects at these red- [4] De Breuck, C., van Breugel, W., Röttger- in TN J1338–1942, only the second shifts which are not out-shined by ing, H., & Miley, G.2000, A&AS, submit- time this is reported in the spectrum of their central AGN like in quasars. Using ted. a z > 4 radio galaxy. The continuum, the high sensitivity of the VLT instru- [5] Dey, A. 1999, in 'The Most Distant Radio shown in the bottom panel of Figure 1, ments, we will be able to study at the Galaxies', ed. H. Röttgering, P. Best M. α same time the origin and evolution of Lehnert (Amsterdam: KNAW), p. 19. shows a discontinuity across the Ly [6] Kauffmann, G., Colberg, J.M., Diaferio, A., line, and a tentative detection of the their giant luminous halos or ionised & White, S. D. M. 1999, MNRAS, 303, 188. Lyman limit. The flux deficit bluewards gas, and the neutral HI gas surrounding [7] Lilly, S. J. & Longair, M., 1984, MNRAS, of Lyα is interpreted as HI absorp- them. 211, 833. tion along the cosmological line of sight, [8] Oke, J.B., & Korycanski, D.G., 1982, ApJ, and is described by the parameter Acknowledgements. The work by 255, 11. D = C.D.B., W.v.B., D.M. and S.A.S. at [9] Pentericci, L. et al. 1999, A&A, 341, 329. A IGPP/LLNL was performed under the [10] Schneider, D. P., Schmidt, M, & Gunn, (Oke & Korycanski 1982). For TN auspices of the US Department of J.E. 1991, AJ, 101, 2004. [11] Spinrad, H., Dey, A., & Graham, J.R. J1338–1942, we measure DA = 0.37 ± Energy under contract W-7405-ENG- 48. D.M. is also supported by Fondecyt 1995, ApJ, 438, 51. 0.1, similar to the DA = 0.45 ± 0.1 [12] van Breugel, W., Stanford, S. A., Spinrad, measured for the z = 4.25 radio galaxy grant No. 01990440 and DIPUC. This H., Stern, D., & Graham, J.R. 1998, ApJ, 8C 1435+64 (Spinrad et al. 1995). Both work is supported in part by the 502, 614. these DA values for z > 4 radio Formation and Evolution of Galaxies [13] van Breugel, W., De Breuck, C., Stanford, galaxies are significantly lower than network set up by the European S. A., Stern, D., Röttgering, H., & Miley, the DA values found for quasars at Commission under contract ERB G. 1999, ApJ, 518, 61. these redshifts (Schneider et al. 1991). FMRX-CT96-086 of its TMR pro- [14] van Ojik, R., Röttgering, H., Miley, G., & Although based on only two measure- gramme. Hunstead, R. W. 1997, A&A, 317, 358.

53 Weighing Young Galaxies – an Occasional Observer Goes to Paranal S. WHITE, Max-Planck Institute for Astrophysics, Garching

“The VLT is Europe’s great leap for- network. Californian graduate students and in airport lounges. The ESO guest ward, heralded as a new window on the now regularly present theses full of 10- house in Santiago, seemingly stuck in a distant universe. Surely we can think of metre data, and Nicole’s contained Keck genteel time warp between encroaching some joint projects that will turn the rotation curves for 15 galaxies out to red- high-rise apartments, provided a wel- Americans green!” Trying to rally the shift 0.8. Data for fifty more were ru- come rest, but was spoiled somewhat by troops at the annual meeting of our EC- moured to be in the pipeline. Some dis- the need to get up long before breakfast funded “Galaxy Formation” network cussion showed that ISAAC could get Hα in order to catch a taxi to catch a plane forced me to think about where the VLT rotation curves for her most distant Keck to catch a taxi to catch the once-a-day pay-off might really come. A talk from galaxies (observed in [OII] 3727 on ESO bus to Paranal. The bus itself, com- Alan Moorwood provided some valu- Hawaii) and, more interestingly, could ex- plete with green baize curtains, reclining able ideas. The first efficient near-IR tend the accessible redshift range by at seats and movie videos, was a surreal but spectrograph on an 8-metre telescope least a factor of two. All we had to do was surprisingly comfortable way to end the could detect Hα past redshift 2 and [OII] to find suitable candidates and to bang odyssey. It dumped us under a clear blue 3727 to redshift 5. How about getting away for a few hours under the famous sky in front of a small city of temporary kinematics for distant galaxies like Paranal observing conditions. A propos- buildings which appeared to have been those in the Hubble Deep Fields or the al was duly dispatched suggesting a pi- built on the Moon. Not Ithaca, perhaps, infamous “Steidel” objects? This would lot study in the Hubble Deep Field South. but it turned out to be too deceptively lux- have to clarify their relation to nearer, We were elated when the Observing Pro- urious to be considered Spartan. dearer, but (perhaps) more boring grammes Committee gave us (most of) In the two days before our run started galaxies. Of course, I’d never taken an the time. we learned the novel joys of P2PP, cre- infrared spectrum of anything, but why Travelling to Paranal from Garching ating enough OBs (Observing Blocks) for not start now? makes one appreciate the virtues of re- the more seasoned observers to decide In a new field it’s always important to mote observing. The shops at Paris CDG that we were quite definitely crazy. Our get good help. My chance came with the and in Buenos Aires were interesting but problem was that it isn't easy to decide news that Nicole Vogt was moving to IoA the attractions of neither place compen- how bright a distant object is likely to be in Cambridge, one of the partners in our sated for the long hours spent on planes in the Hα line when all you have is broadband photometry. Furthermore, we had to make do with photometric redshifts for most of our candidates, so we weren’t sure where Hα would fall relative to the night sky. Finally, there just aren’t any “big” spirals beyond redshift 1 in the Hubble Deep Field South; our best candidates were all dwarf or irregular systems. As a result we needed plenty of flexibility to adjust our programme in light of experience at the telescope – and so plenty of OBs. The inevitable drawbacks of a pilot project... Our first night at the telescope we learned a number of things. (1) ESO looks after its VLT observers extremely well. The night assistant at- tended to all telescope-related issues, including guide-star acquisition. We never had to think about such things. The support scientist knew the instrument and its performance inside out and had written all the available quick- look software. As a result he could help us assess our data as soon as they were taken. Finally, the shift manager always showed up whenever things got complicated. (2) The telescope points and tracks like a dream. Our reference stars always arrived within a few seconds of the centre of the field, our offsets all worked to a small fraction of an arcsecond, even over several arcminutes, and the telescope H-band spectrum of the Hα emission line for disk galaxy HDFS–0620, observed with VLT+ISAAC drifted by only a couple of tenths of an in high-resolution mode. The galaxy has a redshift of z = 1.29, placing the Hα emission at λ = arcsecond even after tracking for two 1.5 µm, in a clear region of the H-band IR window of the atmosphere between night-sky emis- hours. sion lines. The gas can be traced across almost 2 arcseconds (15 kiloparsecs) along the ma- (3) The instrument and the conditions jor axis of the galaxy, yielding a rotation curve with a total observed amplitude of order 180 km/s. really can be “as advertised”. The see-

54 ing on the detector rarely got worse than shifts were in error, their emission lay UVES commissioning team, working 0.5 arcsec and there were extended pe- under a sky line, or was just very weak. 20 metres away on the UT2 console, riods when it was 0.3 arcsec. Furthermore We were greatly relieved when our third obtained their first bona fide quasar the long-slit spectra taken by ISAAC sky- target showed strong Hα at z ~ 1.3 with spectrum. It was certainly impressive; subtracted well and had a cosmetic qual- a clear detection of galactic rotation. the Lyα forest could be seen booming ity quite comparable to that of good op- Our second night was equally good through right down to the atmospheric tical spectra. and brought several more detections cut-off. As we boarded the bus the (4) Even with the best equipment un- as well as some further “blanks”. We following afternoon at the start of the der the best conditions, a hard project is finished with a feeling of achievement; long trek home, we felt sure we would still hard. Our first couple of candidates galaxy kinematics really can be studied be returning soon despite the dis- showed no clear detection after a cou- beyond redshift one. This night also tance. ple of hours of integration, leaving us brought an unexpected (and probably uncertain whether their photometric red- illegal) glass of champagne as the ([email protected])

Sunset on Paranal (Photographer: Herbert Zodet).

55 OTHER ASTRONOMICAL NEWS

Science with the Atacama Large Millimetre Array

“Science with the Ata- cama Large Millimetre Array” was the title of an international conference held at the Carnegie In- stitution in Washington, DC, in early October. In order to increase public attention on the ALMA project, NRAO had organised a press briefing on October 7, enabling journalists to meet with scientists and key representatives of the ALMA partnership. The press briefing followed a reception in the US Congress. In addition, ESO, PPARC and NRAO mounted small exhibitions at the Carnegie Institution itself. C. Madsen

56 ANNOUNCEMENTS

The La Silla 2000+ Report Now Available

The Working Group “La Silla 2000+” has completed its work, http://www.eso.org/gen-fac/commit/ls2000pl1.html based on the replies to the questionnaire survey of the Users Com- mittee (see The Messenger No. 95, p. 38). A total of 256 colleagues The Working Group wishes to thank all of those who answered from the ESO countries participated in this survey, a very gratify- the questionnaire. Special thanks go to the ESO Remedy Team, ing turnout. and in particular to Rein Warmels who helped to design the ques- The report of the Working Group, which contains a set of rec- tionnaire and prepare the figures in the Report. ommendations for the future of the La Silla facilities (1999–2006) The report will now be discussed in the relevant ESO committees. has been submitted to the Director General of ESO (June 1999). The final “La Silla 2000+ Report” (with colour figures) is now Birgitta Nordström available on the Web at the following address: Chair, ESO Users Committee and the WG “La Silla 2000+”

The NEON Observing School

The Network of European Observatories in the North (NEON) committee and will have their travel and living expenses paid, if organises an observational Euro summer school sponsored by the they satisfy the EC rules (age limit of 35 years at the time of the European Community. This school is taking over from the former, Euro Summer School). discontinued, ESO/OHP summer school and intends to run on a The first NEON Euro summer school will take place at the Calar yearly basis. Alto Observatory in Spain, from The participating observatories are: Asiago Observatory (Italy), July 10 to 22, 2000 Calar Alto Observatory (Germany-Spain) and Haute-Provence Ob- servatory (France), with additional tutorial assistance from ESO. Instructions and full practical details can be found on the school The purpose of the school is to provide an opportunity to gain Web site when this announcement appears in print. practical observational experience at the telescope, in observa- Applicants are expected to fill in an application form that will be tories with state-of-the-art instrumentation. To this effect, the school available on the Web site, with a curriculum vitae and a descrip- proposes tutorial observations in small groups of 3 students, un- tion of previous observational experience, and to provide a letter der the guidance of an experienced observer, centred around a of recommendation from a senior scientist. small research project and going through all steps of a standard The provisional application deadline is March 31st, 2000. observing programme. Some complementary lectures will be giv- The school Web site is hosted by the en by experts in the field. EAS Web site at: http://www.iap.fr/eas/schools.html The school is open to students working on a PhD thesis in As- Secretary of the school: Mrs Brigitte RABAN at IAP, 98bis, Bd tronomy and which are nationals of a Member State or an Asso- Arago, F-75014 PARIS. ciated State of the European Union. The working language is Eng- lish. Up to eighteen participants will be selected by the organising Michel Dennefeld, Co-ordinator of the NEON school.

PERSONNEL MOVEMENTS DEPARTURES EUROPE (1st October 1999–30th November 1999) BALON, Serge (F), Technical Assistant BRESOLIN, Fabio (I), Fellow International Staff FERRARI, Marc (F), Associate FUCH, Rainer (D), Legal Advisor HILL, Susan (GB), Archive Operator ARRIVALS KORNWEIBEL, Nicholas (AUS), Software Engineer LATSCH, Hedwig (D), Administrative Clerk EUROPE SCODEGGIO, Marco (I), Fellow

BRISTOW, Paul (GB), Science Data Analyst/Programmer CHILE BROWN, Anthony (NL), Fellow DIMMLER, Martin (D), Electronics/Control Engineer SCHUMANN, Erich (D), Catering La Silla ESKDALE, Jane (AUS), Secretary/Group Assistant GAILLARD, Boris (F), Associate GONÇALVES C., Anabela (P), Fellow RITTER, Heiderose (D), Associate Local Staff RIVINIUS, Thomas (D), Fellow SARTORETTI, Paola (I), Scientific Application Developer DEPARTURES TUELLMANN, Ralph (D), Student VAN ECK, Sophie (B), Fellow DIAZ,Jorge, ZAGGIA, Simone (I), Fellow RAMIREZ, Luis, LECAROS, Fernando, CHILE FERNANDEZ DE C., Rodrigo,

ATHREYA, Ramana (IND), Fellow GALLIANO, Emmanuel (F), Student ARRIVAL MARCONI, Gianni (I), Operations Staff Astronomer TAMAI, Roberto (I), Mechanical Engineer TAPIA, Mario, Mechanical Engineer

57 MESSENGER INDEX 1999 (Nos. 95–98)

SUBJECT INDEX

INAUGURATION CEREMONY M. Tarenghi: News from the VLT 97, 1 Pancino, R. Rengelink, R. Slijkhuis, A. OF PARANAL OBSERVATORY E. Ettlinger, P. Giordano, and M. Schneer- Wicenec, S. Zaggia: ESO Imaging Sur- 5 MARCH 1999 mann: Performance of the VLT Mirror Coat- vey: Past Activities and Future Prospects ing Unit 97, 4 98, 36 M. Sarazin and J. Navarrete: Climate Varia- Introduction to the ESO Annual Report 1998 bility and Ground-Based Astronomy: The by Prof. Riccardo Giacconi, Director Gen- VLT DATA FLOW OPERATIONS VLT Site Fights Against La Niña 97, 8 eral of ESO 96, 2 NEWS C. Castillo F. and L. Serrano G.: Analysis of Discourse of Prof. Riccardo Giacconi, Direc- the Anomalous Atmospheric Circulation in P.A. Woudt and D. Silva: The NTT Service Ob- tor General of ESO 96, 2 Northern Chile During 1998 97, 10 serving Programme: On the Efficiency of Discourse of Mr. Grage, President of the ESO M. Ferrari and F. Derie: Variable Curvature Service Observing 95, 18 Council 96, 4 Mirrors 97, 11 D. Baade: NGC 4945 in 3 Colours 95, 19 Discourse of His Excellency the President of F. Derie, E. Brunetto, and M. Ferrari: The VLTI “First Light” of UT2! 95, 24 the Republic of Chile, Don Eduardo Frei Test Siderostats Are Ready for First Light Ruiz-Tagle 96, 5 97, 12 OBSERVERS WITH THE VLT The VLT Opening Symposium 96, 6 J.C. Christou, D. Bonaccini, N. Ageorges, and F. Marchis: Myopic Deconvolution of Adap- First Scientific Results with the VLT in Visitor TELESCOPES AND tive Optics Images 97, 14 and Service Mode 98, 1 INSTRUMENTATION Latest News: “First Light” for VLT High-Res- B. Brandl, W. Brandner, E.K. Grebel, H. Zin- olution Spectrograph UVES 97, 22 necker: VLT/ISAAC and HST/WFPC2 Ob- servations of NGC 3603 98, 46 A. Moorwood, J.-G. Cuby, P. Ballester, P. Bier- A. Glindemann, R. Abuter, F. Carbognani, F. R.P. Kudritzki, R.H. Mendez, J.J. Feldmeier, eichel, J. Brynnel, R. Conzelmann, B. De- Delplancke, F. Derie, A. Gennai,P. Gitton, R. Ciardullo, G.H. Jacoby, K.C. Freeman, labre, N. Devillard, A. Van Dijsseldonk, G. P. Kervella, B. Koehler, S. Lévêque, G. De M. Arnaboldi,M.Capaccioli,O.Gerhard, H.C. Finger, H. Gemperlein, C. Lidman, T. Her- Marchi, S. Menardi, A. Michel, F. Paresce, Ford: From Intracluster Planetary Nebulae lin, G. Huster, J. Knudstrup, J.-L. Lizon, H. T. Phan Duc, M. Schöller, M. Tarenghi, R. to High-Redshift Lyα Emitters 98, 50 Mehrgan, M. Meyer, G. Nicolini, A. Silber, Wilhelm: The VLT – The Observatory of the C. De Breuck, W. Van Breugel, D. Minniti, G. J. Spyromilio, J. Stegmeier: ISAAC at the 21st Century 98, 2 Miley, H. Röttgering, S.A. Stanford, C. Ca- VLT 95, 1 D. Bonaccini, W. Hackenberg, M. Cullum, M. rilli: VLT Spectroscopy of the z = 4.11 Ra- N. Devillard, Y. Jung, J.-G. Cuby: ISAAC Quattri, E. Brunetto, J. Quentin, F. Koch, E. dio Galaxy TN J1338–1942 98, 52 Pipeline Data Reduction 95, 5 Allaert, A. Van Kesteren: Laser Guide Star S. White: Weighing Young Galaxies – an A. Kaufer, O. Stahl, S. Tubbesing, P. Nørre- Facility for the ESO VLT 98, 8 Occasional Observer Goes to Paranal gaard, G. Avila, P. François, L. Pasquini, A. W. Hackenberg, D. Bonaccini, G. Avila: VLT 98, 54 Pizzella: Commissioning FEROS, the New Laser Guide Star Facility Sub-system De- High-resolution Spectrograph at La Silla sign. Part I: Fibre Relay Module 98, 14 95, 8 New Pictures from the VLT 98, 19 REPORTS FROM OBSERVERS N. Ageorges and N. Hubin: Monitoring of J. Spyromilio, A. Wallander, M. Tarenghi: the Atmospheric Sodium above La Silla Commissioning of the Unit Telescopes of B. Stecklum, H.-U. Käufl, A. Richichi: The Lu- 95, 12 the VLT 98, 21 nar Occultation of CW Leo – a Great Finale D. Baade, K. Meisenheimer, O. Iwert, J. Alon- A. Renzini, P. Rosati: The Science Verification for TIMMI 95, 25 so, Th. Augusteijn, J. Beletic, H. Bellemann, Plan for FORS2 and UVES at UT2/Kuyen L. Guzzo, H. Böhringer, P. Schuecker, C.A. W. Benesch, A. Böhm, H. Böhnhardt, J. 98, 24 Collins, S. Schindler, D.M. Neumann, S. De Brewer, S. Deiries, B. Delabre, R. Donald- R. Gilmozzi: The First Six Months of VLT Sci- Grandi, R. Cruddace, G. Chincarini, A.C. son, Ch. Dupuy, P. Franke, R. Gerdes, A. ence Operations 98, 25 Edge, P.A. Shaver, W. Voges: The Reflex Gilliotte, B. Grimm, N. Haddad, G. Hess, F. Carbognani, G. Filippi, P. Sivera: Configu- Cluster Survey: Observing Strategy and G. Ihle, R. Klein, R. Lenzen, J.-L. Lizon, D. ration Management of the Very Large Tele- First Results on Large-Scale Structure Mancini, N. Munch, A. Pizarro, P. Prado, G. scope Control Software 98, 27 95, 27 Rahmer, J. Reyes, F. Richardson, E. Ro- M. Dolensky, M. Albrecht, R. Albrecht, P. J.U. Fynbo, B. Thomsen, P. Møller: NTT bledo, F. Sanchez, A. Silber, P. Sinclaire, Ballester, C. Boarotto, A. Brighton, T. Service Mode Observations of the Ly- R. Wackermann, S. Zaggia: The Wide Field Chavan, G. Chiozzi, A. Disaro, B. Kemp man-Limit Absorber towards Q1205–30 Imager at the 2.2-m MPG/ESO Telescope: 98, 30 95, 32 First Views with a 67-Million-Facette Eye C. Aerts, P. de Cat and L. Eyer: Long-Term THE LA SILLA NEWSPAGE 95, 15 Spectroscopic Monitoring of Pulsating B R. Kurz and P. Shaver: The ALMA Project O.R. Hainaut: News from the NTT 95, 17 Stars: a Tribute to the CAT 96, 23 96, 7 C. Lidman: New SOFI Grisms – NTT and IR H. Lamy and D. Hutsemékers: A Procedure for A First for the VLT: Observations of the Teams 95, 17 Deriving Accurate Linear Polarimetric Mea- Gamma-Ray Burst GRB990510, and Dis- M. Sterzik: A New Control Room for the 3.6- surements 96, 25 covery of Linear Polarisation 96, 11 m Telescope 95, 18 F. Courbin, P. Magain, S. Sohy, C. Lidman, G. A. Renzini: FORS1 and ISAAC Science Ver- O. Hainaut and the NTT Team: News from the Meylan: Deconvolving Spectra of Lensing ification at Antu/UT1 96, 12 NTT 97, 24 Galaxies, QSO Hosts, and More… 97, 26 B. Renzini: UT2/Kueyen: a Stellar Astron- M. Sterzik, U. Weilenmann and the 3.6-m Up- P. Leisy, P. Francois, and P. Fouqué: Emission- omer’s Dream Comes True 96, 13 grade Team: 3.6-m Telescope Control Sys- Line Object Survey in the LMC with the WFI: G. Monnet: VLT Instrumentation Renewal 96, tem Upgrade Completed 97, 25 New Faint Planetary Nebulae 97, 29 15 Centerfold: Satellite image showing the pro- EIS – THE ESO IMAGING OTHER ASTRONOMICAL NEWS posed location of ALMA 96, 15 SURVEY P. Ballester, A. Disaro, D. Dorigo, A. Modigliani, H.W. Duerbeck, D.E. Osterbrock, L.H. Barrera J.A. Pizarro de la Iglesia: The VLT Data A. Renzini, L. Da Costa: The ESO Imaging S., R. Leiva G.: Halfway from La Silla to Quality Control System 96, 19 Survey 98, 33 Paranal – in 1909 95, 34 E. Allaert: The VLT Software Workshop 96, L. Da Costa, S. Arnouts, C . Benoist, E. B. Nordström: The Questionnaire Survey for 22 Deul, R. Hook, Y.-S. Kim, M. Nonino, E. “La Silla 2000+” 95, 38

58 View of Northern Chile 96, 1 Community 96, 31 Translation of the Speech by H.E. the Presi- ANNOUNCEMENTS List of Scientific Preprints (April–June 1999) dent of the Republic of Chile, Don Eduar- 96, 31 do Frei Ruiz-Tagle, at the Inauguration of Catherine Cesarsky – ESO’s Next Director New ESO Proceedings Available 96, 32 the Paranal Observatoy, 5th March 1999 General 95, 38 Personnel Movements 97, 31 96, 28 ESO Studentship Programme 95, 39 List of Scientific Preprints 97, 32 J. Bergeron and U. Grothkopf: Publications in Personnel Movements 95, 39 B. Nordström: The La Silla 2000+ Report Now Refereed Journals Based on Telescope Ob- List of Scientific Preprints 95, 39 Available 98, 57 servations 96, 28 Minor Planet Mariotti 95, 40 M. Dennefeld: The NEON Observing School C. Madsen: ESO at the Hannover Fair 96, 29 Personnel Movements 96, 30 98, 57 C. Madsen: Science with the Atacama Mil- ESO Fellowship Programme 2000 96, 31 Personnel Movements (1st October – 30th No- limetre Array 98, 56 Contributed Software by Observers in the ESO vember 1999) 98, 57

AUTHOR INDEX

A C. De Breuck, W. Van Breugel, D. Minniti, G. O.R. Hainaut: News from the NTT 95, 17 C. Aerts, P. de Cat and L. Eyer: Long-Term Miley, H. Röttgering, S.A. Stanford, C. Ca- O. Hainaut and the NTT Team: News from the Spectroscopic Monitoring of Pulsating B rilli: VLT Spectroscopy of the z = 4.11 Ra- NTT 97, 24 Stars: a Tribute to the CAT 96, 23 dio Galaxy TN J1338–1942 98, 52 K N. Ageorges and N. Hubin: Monitoring of the M. Dennefeld: The NEON Observing School A. Kaufer, O. Stahl, S. Tubbesing, P. Nørre- Atmospheric Sodium above La Silla 95, 12 98, 57 gaard, G. Avila, P. François, L. Pasquini, A. E. Allaert: The VLT Software Workshop F. Derie, E. Brunetto, and M. Ferrari: The VLTI Pizzella: Commissioning FEROS, the New 96, 22 Test Siderostats Are Ready for First Light High-resolution Spectrograph at La Silla 97, 12 95, 8 B N. Devillard, Y. Jung, J.-G. Cuby: ISAAC Pipe- R.P. Kudritzki, R.H. Méndez, J.J. Feldmeier, D. Baade, K. Meisenheimer, O. Iwert, J. Alon- line Data Reduction 95, 5 R. Ciardullo, G.H. Jacoby, K.C. Freeman, so, Th. Augusteijn, J. Beletic, H. Bellemann, M. Dolensky, M. Albrecht, R. Albrecht, P. Bal- M. Arnaboldi, M. Capaccioli, O. Gerhard, W. Benesch, A. Böhm, H. Böhnhardt, J. lester, C. Boarotto, A. Brighton, T. Chavan, H.C. Ford: From Intracluster Planetary Neb- Brewer, S. Deiries, B. Delabre, R. Donald- G. Chiozzi, A. Disaro, B. Kemp 98, 30 ulae to High-Redshift Lyα Emitters 98, 50 son, Ch. Dupuy, P. Franke, R. Gerdes, A. H.W. Duerbeck, D.E. Osterbrock, L.H. Barrera R. Kurz and P. Shaver: The ALMA Project Gilliotte, B. Grimm, N. Haddad, G. Hess, S., R. Leiva G.: Halfway from La Silla to 96, 7 G. Ihle, R. Klein, R. Lenzen, J.-L. Lizon, D. Paranal – in 1909 95, 34 Mancini, N. Munch, A. Pizarro, P. Prado, G. L Rahmer, J. Reyes, F. Richardson, E. Ro- E H. Lamy and D. Hutsemékers: A Procedure for bledo, F. Sanchez, A. Silber, P. Sinclaire, E. Ettlinger, P. Giordano, and M. Schneer- Deriving Accurate Linear Polarimetric Mea- R. Wacker-mann, S. Zaggia: The Wide mann: Performance of the VLT Mirror Coat- surements 96, 25 Field Imager at the 2.2-m MPG/ESO Tele- ing Unit 97, 4 P. Leisy, P. François, and P. Fouqué: Emis- scope: First Views with a 67-Million- sion-Line Object Survey in the LMC with Facette Eye 95, 15 F the WFI: New Faint Planetary Nebulae D. Baade: NGC 4945 in 3 Colours 95, 19 M. Ferrari and F. Derie: Variable Curvature Mir- 97, 29 P. Ballester, A. Disaro, D. Dorigo, A. Modigliani, rors 97, 11 C. Lidman: New SOFI Grisms – NTT and IR J.A. Pizarro de la Iglesia: The VLT Data Discourse of His Excellency the President of Teams 95, 17 Quality Control System 96, 19 the Republic of Chile, Don Eduardo Frei J. Bergeron: The VLT Opening Symposium Ruiz-Tagle 96, 5 96, 6 J.U. Fynbo, B. Thomsen, P. Møller: NTT Ser- M J. Bergeron and U. Grothkopf: Publications in vice Mode Observations of the Lyman-Lim- C. Madsen: ESO at the Hannover Fair Refereed Journals Based on Telescope it Absorber towards Q1205–30 95, 32 96, 29 Observations 96, 28 C. Madsen: Science with the Atacama Mil- D. Bonaccini, W. Hackenberg, M. Cullum, M. G limetre Array 98, 56 Quattri, E. Brunetto, J. Quentin, F. Koch, E. Introduction to the ESO Annual Report 1998 G. Monnet: VLT Instrumentation Renewal Allaert, A. Van Kesteren: Laser Guide Star by Prof. Riccardo Giacconi, Director Gen- 96, 15 Facility for the ESO VLT 98, 8 eral of ESO 96, 2 A. Moorwood, J.-G. Cuby, P. Ballester, P. Bier- B. Brandl, W. Brandner, E.K. Grebel, H. Zin- Discourse of Prof. Riccardo Giacconi, Direc- eichel, J. Brynnel, R. Conzelmann, B. De- necker: VLT/ISAAC and HST/WFPC2 Ob- tor General of ESO 96, 2 labre, N. Devillard, A. Van Dijsseldonk, G. servations of NGC 3603 98, 46 R. Gilmozzi: The First Six Months of VLT Sci- Finger, H. Gemperlein, C. Lidman, T. Her- ence Operations 98, 25 lin, G. Huster, J. Knudstrup, J.-L. Lizon, H. C A. Glindemann, R. Abuter, F. Carbognani, F. Mehrgan, M. Meyer, G. Nicolini, A. Silber, F. Carbognani, G. Filippi, P. Sivera: Configu- Delplancke, F. Derie, A. Gennai, P. Gitton, J. Spyromilio, J. Stegmeier: ISAAC at the ration Management of the Very Large Tele- P. Kervella, B. Koehler, S. Lévêque, G. De VLT 95, 1 scope Control Software 98, 27 Marchi, S. Menardi, A. Michel, F. Paresce, C. Castillo F. and L. Serrano G.: Analysis of T. Phan Duc, M. Schöller, M. Tarenghi, R. N the Anomalous Atmospheric Circulation in Wilhelm: The VLT – The Observatory of the B. Nordström: The Questionnaire Survey for Northern Chile During 1998 97, 10 21st Century 98, 2 “La Silla 2000+” 95, 38 J.C. Christou, D. Bonaccini, N. Ageorges, and Discourse of Mr. Grage, President of the ESO B. Nordström: The La Silla 2000+ Report Now F. Marchis: Myopic Deconvolution of Adap- Council 96, 4 Available 98, 57 tive Optics Images 97, 14 L. Guzzo, H. Böhringer, P. Schuecker, C.A. F. Courbin, P. Magain, S. Sohy, C. Lidman, G. Collins, S. Schindler, D.M. Neumann, S. De R Meylan: Deconvolving Spectra of Lensing Grandi, R. Cruddace, G. Chincarini, A.C. A.Renzini: FORS1 and ISAAC Science Veri- Galaxies, QSO Hosts, and More... 97, 26 Edge, P.A. Shaver, W. Voges: The Reflex fication at Antu/UT1 96, 12 Cluster Survey: Observing Strategy and First A. Renzini: UT2/Kueyen: a Stellar Astron- D Results on Large-Scale Structure 95, 27 omer’s Dream Comes True 96, 13 L. Da Costa, S. Arnouts, C. Benoist, E. Deul, A. Renzini, P. Rosati: The Science Verification R. Hook, Y.-S. Kim, M. Nonino, E. Panci- H Plan for FORS2 and UVES at UT2/Kuyen no, R. Rengelink, R. Slijkhuis, A. Wicenec, W. Hackenberg, D. Bonaccini, G. Avila: VLT 98, 24 S. Zaggia: ESO Imaging Survey: Past Ac- Laser Guide Star Facility Sub-system De- A. Renzini, L. Da Costa: The ESO Imaging tivities and Future Prospects 98, 36 sign. Part I: Fibre Relay Module 98, 14 Survey 98, 33

59 M. Sterzik: A New Control Room for the ESO, the European Southern Observatory, S 3.6-m Telescope 95, 18 was created in 1962 to “… establish and op- M. Sarazin and J. Navarrete: Climate Variability and Ground-Based Astronomy: The VLT erate an astronomical observatory in the T southern hemisphere, equipped with pow- Site Fights Against La Niña 97, 8 J. Spyromilio, A. Wallander, M. Tarenghi: Com- M. Tarenghi: News from the VLT 97, 1 erful instruments, with the aim of furthering missioning of the Unit Telescopes of the VLT and organising collaboration in astronomy 98, 21 …” It is supported by eight countries: Bel- W B. Stecklum, H.-U. Käufl, A. Richichi: The Lu- S. White: Weighing Young Galaxies – an gium, Denmark, France, Germany, Italy, the nar Occultation of CW Leo – a Great Finale Occasional Observer Goes to Paranal Netherlands, Sweden and Switzerland. ESO for TIMMI 95, 25 98, 54 operates at two sites. It operates the La Sil- M. Sterzik, U. Weilenmann and the 3.6-m P.A. Woudt and D. Silva: The NTT Service Ob- la observatory in the Atacama desert, 600 Upgrade Team: 3.6-m Telescope Control serving Programme: On the Efficiency of km north of Santiago de Chile, at 2,400 m System Upgrade Completed 97, 25 Service Observing 95, 18 altitude, where several optical telescopes with diameters up to 3.6 m and a 15-m sub- millimetre radio telescope (SEST) are now in operation. In addition, ESO is in the process of building the Very Large Telescope (VLT) on Paranal, a 2,600 m high mountain Contents approximately 130 km south of Anto- fagasta, in the driest part of the Atacama First Scientific Results with the VLT in Visitor and Service Modes ...... 1 desert. The VLT consists of four 8.2-metre and three 1.8-metre telescopes. These telescopes can also be used in combination as TELESCOPES AND INSTRUMENTATION a giant interferometer (VLTI). The first 8.2- metre telescope (called ANTU) is since April A. Glindemann, R. Abuter, F. Carbognani, F. Delplancke, F. Derie, A. Gennai, 1999 in regular operation, and also the sec- P. Gitton, P. Kervella, B. Koehler, S. Lévêque, G. De Marchi, S. Menardi, ond one (KUEYEN) has already delivered A. Michel, F. Paresce, T. Phan Duc, M. Schöller, M. Tarenghi, R. Wilhelm: pictures of excellent quality. Over 1200 pro- The VLTI – The Observatory of the 21st Century ...... 2 posals are made each year for the use of D. Bonaccini, W. Hackenberg, M. Cullum, M. Quattri, E. Brunetto, J. Quentin, the ESO telescopes. The ESO Headquar- F. Koch, E. Allaert, A. Van Kesteren: Laser Guide Star Facility for the ESO ters are located in Garching, near Munich, VLT ...... 8 Germany. This is the scientific, technical and W. Hackenberg, D. Bonaccini, G. Avila: VLT Laser Guide Star Facility Sub- administrative centre of ESO where tech- system Design. Part I: Fibre Relay Module ...... 14 nical development programmes are carried New Pictures from the VLT ...... 19 out to provide the La Silla and Paranal observatories with the most advanced instru- J. Spyromilio, A. Wallander, M. Tarenghi: Commissioning of the Unit ments. There are also extensive astro- Telescopes of the VLT ...... 21 nomical data facilities. In Europe ESO em- A. Renzini, P. Rosati: The Science Verification Plan for FORS2 and UVES at ploys about 200 international staff members, UT2/Kuyen ...... 24 Fellows and Associates; in Chile about 70 R. Gilmozzi: The First Six Months of VLT Science Operations ...... 25 and, in addition, about 130 local staff mem- F. Carbognani, G. Filippi, P. Sivera: Configuration Management of the Very bers. Large Telescope Control Software ...... 27 M. Dolensky, M. Albrecht, R. Albrecht, P. Ballester, C. Boarotto, A. Brighton, The ESO MESSENGER is published four T. Chavan, G. Chiozzi, A. Disarò, B. Kemp: Java for Astronomy: Software times a year: normally in March, June, Sep- Development at ESO/ST-ECF ...... 30 tember and December. ESO also publish- es Conference Proceedings, Preprints, EIS – THE ESO IMAGING SURVEY Technical Notes and other material con- nected to its activities. Press Releases in- A. Renzini, L. Da Costa: The ESO Public Imaging Survey ...... 33 form the media about particular events. For L. Da Costa, S. Arnouts, C. Benoist, E. Deul, R. Hook, Y.-S. Kim, M. Nonino, further information, contact the ESO Edu- E. Pancino, R. Rengelink, R. Slijkhuis, A. Wicenec, S. Zaggia: cation and Public Relations Department at ESO Imaging Survey: Past Activities and Future Prospects ...... 36 the following address: OBSERVERS WITH THE VLT EUROPEAN SOUTHERN OBSERVATORY B. Brandl, W. Brandner, E.K. Grebel, H. Zinnecker: VLT/ISAAC and Karl-Schwarzschild-Str. 2 HST/WFPC2 Observations of NGC 3603 ...... 46 D-85748 Garching bei München R.P. Kudritzki, R.H. Mendez, J.J. Feldmeier, R. Ciardullo, G.H. Jacoby, Germany K.C. Freeman, M. Arnaboldi, M. Capaccioli, O. Gerhard, H.C. Ford: From Tel. (089) 320 06-0 α Telefax (089) 3202362 Intracluster Planetary Nebulae to High-Redshift Ly Emitters ...... 50 [email protected] (internet) C. De Breuck, W. Van Breugel, D. Minniti, G. Miley, H. Röttgering, URL: http://www.eso.org S.A. Stanford, C. Carilli: VLT Spectroscopy of the z = 4.11 Radio Galaxy TN J1338–1942 ...... 52 S. White: Weighing Young Galaxies – an Occasional Observer Goes to The ESO Messenger: Paranal ...... 54 Editor: Marie-Hélène Demoulin Technical editor: Kurt Kjär OTHER ASTRONOMICAL NEWS

Printed by C. Madsen: Science with the Atacama Millimetre Array ...... 56 J. Gotteswinter GmbH Buch- und Offsetdruck ANNOUNCEMENTS Joseph-Dollinger-Bogen 22 D-80807 München B. Nordström: The La Silla 2000+ Report Now Available ...... 57 Germany M. Dennefeld: The NEON Observing School ...... 57 Personnel Movements (1st October – 30th November 1999) ...... 57 ISSN 0722-6691 MESSENGER INDEX 1999 (Nos. 95–98) ...... 58