1

No. 94 – December 1998

1. Introduction

1.1 The Project

FORS, the FOcal Reducer/low dis-persion Spectrograph, was the first VLTinstrument to be designed and built out-side ESO. Following a Call for Proposalsin 1990, the contract to realise the pro-ject was awarded in 1991 to a consortiumof three German astronomical institutes(Landessternwarte Heidelberg and theUniversity Observatories of Göttingenand Munich). Due to its large variety ofobserving modes, a heavy demand for ob-serving time was expected for the envis-aged instrument. Therefore, ESO decid-ed from the beginning to order two near-ly identical instruments from the consor-tium. An account of the start of the pro-ject is given in [1]. In April 1992 thePreliminary Design Review was held fol-lowed by the Final Design Review inFebruary 1994. Immediately afterwards,the consortium started to transform theapproved design into hardware. The per-formance of the instrument and its com-ponents was continuously tested in thelaboratory during the different stages of

Successful Commissioning of FORS1 – the First Optical Instrument on the VLTI. APPENZELLER1, K. FRICKE2, W. FÜRTIG1, W. GÄSSLER 3, R. HÄFNER 3, R. HARKE 2, H.-J. HESS 3, W. HUMMEL3, P. JÜRGENS2, R.-P. KUDRITZKI 3, K.-H. MANTEL3, W. MEISL3, B. MUSCHIELOK 3, H. NICKLAS 2, G. RUPPRECHT 4, W. SEIFERT 1, O. STAHL1,T. SZEIFERT 1, K. TARANTIK 3

1Landessternwarte Heidelberg, 2Universitäts-Sternwarte Göttingen, 3Universitäts-Sternwarte München,4ESO Garching

Figure 1: FORS1 installed on UT1. The instrument with the electronics boxes is well visible aswell as the tripod on the M1 cell that carries the cable hose. This cable hose (a flexible metalhose) contains all electric power, LAN and cooling water lines for FORS.

manufacturing and assembling. In No-vember 1996 extensive tests of the inte-grated FORS1 instrument were started inan Integration Facility of the DeutscheForschungsanstalt für Luft- und Raum-fahrt (DLR) in Oberpfaffenhofen nearMunich. A telescope and star simulatorallowed to tilt and rotate the instrumentand to simulate stars in the field of view.Electro-mechanical functions, image mo-tion due to flexure, optical performance,the calibration units and the instrument-related software were checked for theirconformity with the specifications givenby ESO and were optimised if deemednecessary. For details see [7] and [8].

Within the consortium the tasks weredistributed as follows:

• Heidelberg: optics, data reductionsoftware, telescope and star simulator

• Göttingen: mechanics, auxiliary de-vices

• Munich: instrument-related software,electronics, project management

The detector system, based on a SITe2048 × 2048 CCD with 24-µm pixel size,was developed by ESO.

The manufacturing of the twin instru-ment FORS2 will be completed by the endof 1998. After the performance tests,which will take about half a year, FORS2will be shipped to Paranal in summer1999. It will be installed on VLT UT2 atthe end of 1999 and released for obser-vation in March 2000.

1.2 Design Parameters

FORS has been designed as an all-dioptric Cassegrain instrument for thewavelength range from 330 nm to 1100nm and it will cover the following ob-serving modes:

• Direct imaging with two image scales(0.2 and 0.1 arcsec/pixel, correspondingfields of view of 6.8′ × 6.8′ and 3.4′ × 3.4′,respectively)

• Low-dispersion grism spectroscopy(resolutions between 200–1800 for 1 arc-sec slit width, reciprocal dispersion44–228 Å/mm)

– Multi-object spectroscopy (MOS) ofup to 19 targets with individually ad-justable slits

– Long-slit spectroscopy (mask with 9slits with fixed widths between 0.3 and 2.5arcsec)

• Imaging polarimetry and spectro-po-larimetry (linear and circular; FORS1only)

• Echelle spectroscopy (FORS2 only),with resolutions up to about 5000 for1 arcsec slit width, reciprocal dispersion15–40 Å/mm

• Multi-object spectroscopy with slitmasks (FORS2 only), for survey-typeprogrammes where more than 19 objectsare observed or special slit shapes areneeded.

A more comprehensive description ofthe instrument performance is given in [2],[3] and [4]. The grisms used in FORS aredescribed in [5].The main scientific objective of FORS wasto extend ground-based spectroscopyand photometry to significantly fainter ob-jects than could be reached so far.Among the tasks of FORS will be thequantitative analysis of the properties ofgalaxies at distances up to 10 billion light-years and beyond. Thus it will becomepossible to obtain information on the uni-verse in its early phases resulting in fun-damental insights into the creation anddevelopment of galaxies. Investigationsof clusters of galaxies and of single starsin galaxies may shed light on the natureof the dark matter. Spectroscopy of lu-minous stars, supernovae and planetarynebulae in remote galaxies will allow usto obtain a better knowledge of the ex-pansion of the universe and will provideinformation on the origin of chemical el-ements outside our own Galaxy. Even in

our Galaxy new discoveries concerninge.g. the late phases of stellar evolution orthe creation of stars and planetary sys-tems are to be expected (see also [6]).

2. The Final Steps

By the beginning of 1998 the FORSproject started to home in on the finalround. The instrument had already beentested extensively in the DLR facility.In January ESO’s Optical Detector Teamdelivered the final scientific CCD system,together with one of the first copies of thenewly developed FIERA CCD controllers.

Another round of tests followed whichalso served as the “First VerificationTest” specified in the contract to be donebefore shipping the instrument to Chile.During these tests particular attention waspaid to proving that the image quality inall observing modes conformed to thespecifications. During another extensiveset of measurements we checked for im-age motion due to instrument flexure(Fig. 2). This is particularly important forFORS as a Cassegrain instrument that issubjected to wide variations in spatial ori-entation during tracking. The results arepublished in [8] and fully comply with thefinite element calculations done during thedesign phase. Other tests included thesetting accuracy of the slitlets for themulti-object spectroscopy, which is cru-cial for the successful use of that mode.Moreover, we tested already in Europethe complete control system, from theworkstation software through the localarea network connection to the softwarerunning on the three instrument local con-trol units ( “board computers” ) and theirco-operation with the FIERA detectorcontroller. In addition, the interface to theTelescope Co-ordination Software (TCS)was tested with the TCS simulator atESO/Garching. All tests were passed suc-cessfully, and by the middle of JuneFORS was removed from the telescopesimulator to be packed for shipping.

It had been decided to transport FORSby air to Chile. This meant first a trans-port by truck from Oberpfaffenhofen toFrankfurt airport and from there by air-plane to Santiago de Chile. The packingtherefore had to take into account the spe-cific requirements of this multiple trans-port: The instrument was taken apart asfar as necessary to allow safe packing indedicated boxes. The boxes for the ma-jor instrument sections consisted of stur-dy frames on which wooden boxes weresitting, damped by specially dimensionedcoiled steel springs. All other componentslike drive units, electronic racks, filters,grisms, etc. were packed in aluminiumcontainers. Altogether the shipment con-sisted of 22 boxes. After the transport hadarrived in Santiago they were loaded ontotwo climatised trucks with air suspensionfor the final 1200-km journey to the VLTObservatory on Paranal. Thanks to thevery professional way this was handledby the truck drivers, also this 2-day trippassed without problems – under the

2

Figure 2: FORS1 on the telescope simulator during flexure tests in the integration hall of DLRat Oberpfaffenhofen near Munich.

watchful eyes of the accompanying pro-ject manager.

In the meantime, the preparations onParanal for the re-assembly of FORS1were nearly complete: erection of a spe-cially designed integration stand, settingup of a local-area network for our work-station and many other little things whichwould allow us to start integration of theinstrument immediately after its arrival onthe mountain.

Re-assembly in the Auxiliary TelescopeHall (ATH, Fig. 3) in the Paranal basecamp went very smoothly. The instrumentsections were unpacked, transport locksand dampers removed, and the main op-tics inserted. In parallel the electronicswere activated. Finally, all instrumentsections were put together on the inte-gration flange and the detector system at-tached. Another round of tests involvingfunctional checks of all motors and seriesof screen flats and pinhole exposures fol-lowed which proved that we were reallyready for installation on the telescope.

In the morning of September 10, theATH crane was used to put FORS1 (wellpacked in plastic foil to protect it againstdust during the following 3-km journey)onto the loading area of the large Paranalair suspension truck. It is a strange feel-ing to watch the work of the past 9 yearshover 5 metres above the ground, sup-ported only by a couple of thin wires!

Very slowly and carefully, the truckmoved up to the top of the mountain, ac-companied by an escort of cars forcingall other traffic out of the way, and somecolleagues constantly having an eye onthe truck with its precious load (Fig. 4).After nearly one hour, the convoy arrivedin front of UT1. The following operationproved to be the most time consuming ofthe whole transport and installation: get-

out problems. Next was the loading of allauxiliary optics (filters, grisms, Wollastonprism) in the corresponding openings ofthe filter and grism wheels.

3. At Last: First Light!

On the evening of September 15 finallyeverything was ready: The sky overParanal was clear, the enclosure wasopened, the telescope pointed to a pho-tometric standard star field close tozenith, the image analysis was performedby the active optics system and the in-strument shutter opened for a 10-secondexposure through the Bessell V filter. Theexcitement in the control room was grow-ing while the image appeared on the RealTime Display station: the stars lookedneatly small and round – we had suc-cessful First Light! The camera focus hadbeen pre-set to the value calculated ac-cording to the current temperature, andas the following series of exposuresshowed this was exactly the right setting.

The rest of this night and the subse-quent ones were used to conduct aquick series of observations in all in-strument modes to confirm that the over-all performance was as expected. In thefourth night a thorough verification pro-gramme began during which we made ex-posures in all observing modes, with allavailable filters and grisms. Altogetherabout 20 Gb of data were taken duringthe 21 nights of the first commissioningperiod. Some of these data were reducedin a preliminary way and made public im-mediately (see http://www.eso.org/out-reach/press-rel/pr-1998/pr-14-98.html andhttp://www.eso.org/outreach/press-rel/pr-1998/phot-38-98.html); now work is go-ing on with the remainder to prove thatthe instrument indeed fulfils the specifi-cations. A preliminary assessment is giv-en in the following section.

3

ting FORS1 and the carriage down fromthe transport truck using the huge M1 celllifting platform. The carriage with FORS1on it was then lifted through the trap dooropening onto the telescope fork base withthe enclosure crane and soon afterwardsFORS1 was mounted at the UT1 Casse-grain flange! Finally the CCD detector inits cryostat was attached at the FORS1bottom end as well as the tube tripod car-rying the cable hose with the electricalpower, LAN and cooling water lines. Thenext day saw the first test of the completeFORS1 system in the telescope envi-ronment (Fig. 1) which also went with-

Figure 3: Re-assembly in the Auxiliary Telescope Hall on Paranal with FORS1 on the integra-tion stand in the background, the control cabin at right and the Cassegrain carriage (foregroundleft) which is used for all major transport and handling operations.

Figure 4: FORS1 on its way to the telescope.

4

4. Preliminary Results

During the first commissioning phase,all observing modes of FORS1 weresystematically tested with astronomicaltargets. The commissioning tests span avery wide range from photometric cali-bration through astrometric tests to ex-tensive polarimetric observations. Thesecond main topic during this phase wasthe proper interaction between telescope,instrument and the software user inter-faces (see below). Most optical testswere done both in high and standard res-olution mode and for a large number ofoptical components. Although theserepetitive tests dominated our work wealso obtained some eye-catching sou-venirs, e.g. observing the splendid face-on spiral galaxy NGC 1232 and theDumbbell nebula M27 (Fig. 5), both ofwhich fit very well in the field of view instandard resolution mode.

After attaching FORS1 to the tele-scope, the standard maintenance pro-cedures for preparing the instrument forastronomical work were executed. Theyinclude an automatic check of allelectromechanical functions, insertion ofoptical components as well as the align-ment of CCD, grisms and Wollastonprism. At the end of this stage the in-strument was ready for the “first astro-nomical light”. This first light was usedfor some initialisation procedures liketelescope/instrument focusing or deter-

mination of the angle between the CCDrows/columns and the telescope axes.After this was done, the instrument wasready for astronomical work. The firstobservations were performed with anengineering Instrument Control Panelwhich allows the execution of singleexposures. Later on, FORS templates asexecuted by BOB (the broker of obser-vation blocks) were used.

Maintenance and Instrument ControlSoftware were working perfectly almostwithout any modifications.

The Observer Support Software is aGraphical User Interface based on ESOSKYCAT, which will be used for thepreparation of observations in all ob-serving modes. It allows e.g. the prepa-ration of the MOS set-up, i.e. it specifiesthe telescope pointing position, the tele-scope rotator position and assigns targetsand MOS slits. The astronomer will usethis tool in advance for preparation of theobservations.

The FORS templates provide supportfor the following observing modes: imag-ing (IMG), imaging polarimetry (IPOL),long-slit spectroscopy (LSS), multi-objectspectroscopy (MOS) and multi-objectspectropolarimetry (PMOS). For eachmode two templates exist: one for inter-nal calibration (screen flat and wavelengthcalibration) and the other for science ob-servation and external calibration (sky flatand standard stars). Templates necessaryfor all modes are collected in the categoryALL (e.g. dark, focus or acquisition tem-plates). The templates are built in such away that all features of FORS can be usedin their full variety.

Observation Software is triggered bythe FORS templates to perform the set-up of instrument, detector and telescopeand to execute the exposure. It writes alsothe header information in the output FITSfile. In interaction with FORS templates,this module is also responsible for thealignment of targets and slits for spec-troscopic observations.

One of the main characteristics ofFORS is the very high stiffness of the in-strument. Any motion of the image on thedetector would degrade the instrumentperformance.

Therefore, a passive flexure compen-sation scheme was part of the FORS de-sign. The passive system compensatesany motion of the collimating optics by tilt-ing the camera.

Figure 5: Colour image of the Dumbbell planetary nebula (M 27), obtained on September 28,1998. This is a three-colour composite based on two interference ([OIII] at 501 nm and 6 nmFWHM – 5 min exposure time; H-alpha at 656 nm and 6 nm FWHM – 5 min) and one broad-band (Bessell B at 429 nm and 88 nm FWHM; 30 sec) filter images. North is up; East is left.

The TelescopeControl Softwarewith the real tele-scope behind it andreal sky imagesconstituted two newinterfaces for theFORS Software.They introducedmodifications in itsuppermost layers,i.e. in the ObserverSupport Software(OSS), in the FORStemplates and in theObservation Soft-ware (OS). The re-maining softwaremodules i.e. the

Figure 6: Profile of astellar point spreadfunction in standardresolution mode.

Tests with our telescope simulatorconfirmed that the internal image motionfrom the focal plane of the telescope tothe focal plane of FORS is smaller thana quarter of a pixel during long exposuresof one hour in standard and two hours inhigh-resolution mode.

External flexure and motions of FORSrelative to the guide probe and the guid-ing accuracy were tested with a long timeseries observation of an empty field,which showed that the external flexureand the guiding can be kept within a frac-tion of a pixel during several hours of ob-serving time and thus complies with thecontractual specifications.

In direct imaging, the image qualitymeasured on the detector was alwayslimited by the outside seeing. Most ofthe time it was better than the valuesindicated by the DIMM seeing moni-tor and always compatible with the mea-surements by the adapter guide probe.The best FWHM values achieved rangefrom 0.48″ for a 4-min. exposure in theU band to 0.35″ in the I band (1 min.). Noimage degradation due to the FORS1 op-tics could be detected over the wholefield of view. The variation of the stellarFWHM over the field is less than 3% withboth collimators. A composite PSF pro-file for the standard resolution collima-tor is shown in Figure 6. The radius ofthe PSF 10 mag below the peak bright-

ness is about 4″, an excellent value.The characteristics of all grisms were

analysed in long-slit and multi-objectmode using test observations of variousdifferent types of targets. The measuredspectral resolutions were found to matchvery well the theoretically calculated val-ues for two-pixel sampling. As an exam-ple we present in Figure 7 a section of thespectrum of the distant quasar Q0103260(visual magnitude V = 18.8, redshift z =3.36). This spectrum is based on a 30-minute FORS exposure taken with a1.0″ slit and the 600B grism (600 lines/mmblazed in the blue spectral range). The

prominent emission line at 530 nm is theLyman-α resonance line of atomic hy-drogen, redshifted from 122 nm to 530 nm.To the left of the Lyman-α emission linewe see clearly the “Lyman-α forest”formed by numerous Lyman-α absorptionlines of hydrogen clouds of lower redshiftlocated between the quasar and us.Q0103-260 is a particularly interestingobject for studying the Lyman forest sincethis quasar is in a region close to thesouth galactic pole with its exceptionallylow density of foreground stars andforeground galaxies. Therefore, it will bepossible to obtain with FORS on the VLTvery deep images and spectra of faint verydistant objects in the field around thisquasar in order to search for objects thatmay be connected to the hydrogen ab-sorption features in the spectrum of thequasar.

Several multi-object spectra were alsotaken to test the functionality of theobject acquisition and the instrumen-tal behaviour in the MOS mode. In Fig-ure 8 the spectra of 19 stars (including sev-eral Be stars) in the open star clusterNGC330 in the Small Magellanic Cloudare shown.

Imaging polarimetry was done for asample of unpolarised and polarised stan-dard stars, both for linear and circular po-larimetry. Systematic errors due to the factthat our retarder plates are built as mo-saics are less than 2 10–4. Also linear andcircular spectropolarimetry was done tocalibrate the angular zero point of the half-wave plate as a function of wavelengthand to evaluate possible cross-talk be-tween the Stokes parameters. As an ex-ample, the circular polarisation of the14.8-magnitude magnetic white dwarfGD229 is shown in Figure 9. The sharpfeatures in the polarisation spectrumaround 420 nm, 530 nm, 630 nm and 710nm correspond also to broad ‘absorption’features in the flux spectrum. Most of thosecan be explained by stationary Zeemantransitions in a magnetic field of 25 to 60MG. The very low noise of 0.15% wasreached with two exposures of 10 minutes.At this level the noise is in good agreementwith the photon noise and no systematiceffects appear.

As pointed out above, FORS wasdeveloped mainly to do spectroscopy

5

Figure 7: A spectrum of the quasar Q0103-260. The quasar shows a prominent Lyman alphaforest.

Figure 8: Multi object spectroscopy of stars located in the open cluster NGC330 of the SmallMagellanic Cloud, exposure time 6 minutes. The spectra appear as bright lines spanning thefull field in horizontal direction. The shorter, bright vertical lines are spectral emission lines orig-inating in the terrestrial atmosphere (airglow); they show the extent of the individual slits. Notethat in some slits more than one star spectrum has been registered, thus further increasing theobserving efficiency.

on very faint objects. Therefore, we madespectroscopic test observations of dis-tant galaxies up to 25 magnitude inR and I. Moreover, we successfully ob-tained spectra of faint gravitational arcsaround two galaxy clusters. The reduc-tion and evaluation of these data is stillin progress.

5. A Bright Future Ahead

The FORS schedule foresees a sec-ond commissioning period after the as-sessment of the data from the first com-missioning; after this the instrument is of-ficially handed over to ESO who performsone week of Science Verification inJanuary 1999. The beginning of Ob-serving Period 63 on April 1, 1999 marksthe start of regular observations of FORSon UT1 and the arrival of the first VisitingAstronomers on Paranal. By that time alsoa core team of telescope and instrumentoperators will have taken up duty to car-ry out the approved observing pro-grammes. 182 proposals have been sub-mitted for observations with FORS1 aloneduring its first 6 months of service – proofof trust that the instrument will live up tothe expectations! And the extrapolationof our experience gained during the com-missioning truly lets us expect thatFORS1 on the VLT will be a world-classfacility for the exploration of the veryfaintest objects.

6. Acknowledgements

The FORS project would not havebeen possible without the contributions ofnumerous people in the VLT InstrumentConsortium and in ESO. We would liketo thank all those who have contributedto this ambitious project, especially themembers of the workshops of the partic-ipating observatories. The project wouldnot have been successful without a co-operation between the VIC and ESOteams which was, through the entire du-ration, dominated by the common goal ofdelivering a superb instrument to the as-tronomical community of ESO’s membercountries and characterised by an ex-cellent personal relationship between allteam members.

Our special thanks are due to our for-mer collaborators in Heidelberg (C. Hart-lieb, S. Möhler, R. Östreicher, L. Schäff-ner), Göttingen (F. Degenhardt, K.-H.Duensing, U. Duensing, S. Gong, T.Nguyen, R. Pick, V. Radisch, H. Ren-ziehausen, M. Ronnenberg., T. Töteberg,W. Wellem) and Munich (H. Böhnhardt,H. Geus, A. Hebenstreit, S. Kiesewetter-Köbinger, W. König, W. Mitsch, F. Mitter-meier, M. Roth, P. Well).

The valuable contributions by ESOstaff during the FORS development phase(A. Balestra, J. Beletic, C. Cumani, S.Deiries, B. Delabre, S. D’Odorico, R. Don-aldson, R. Dorn, G. Filippi, R. Gilmozzi,G. Hess, O. Iwert, H. Kotzlowski, M.

Kraus, J.-L. Lizon, G. Monnet, W. Nees,R. Reiss, G. Raffi, R. Warmels, G. Wie-land) and the installation and commis-sioning phase (H. Böhnhardt, J. Brynnel,M. Cullum, F. Franza, H. Gemperlein, P.Gray, G. Ihle, I. Osorio, P. Sansgasset,J. Spyromilio, A. van Kesteren, A. Wal-lander, K. Wirenstrand) are gratefullyacknowledged.

The Instrument Science Team (S.Cristiani, S. di Serego Alighieri, Y. Mellier,P. Shaver, J. Surdej) gave many con-structive recommendations.

The FORS test team benefited from theDLR management (H. Holzach) and thelocal workshop. Last but not least: TheFORS project could not have been re-alised without the financial support by theGerman Federal Ministry of Education,Science, Research and Technology(BMBF grants 05 3HD50A, 053 GO10Aand 053 MU104).

References

1. Appenzeller, I., Rupprecht, G.: FORS, TheFocal Reducer for the VLT, The MessengerNo. 67, p. 18, 1992.

2. Mitsch, W., Rupprecht, G., Seifert, W.,Nicklas, H., Kiesewetter, S.: Versatile mul-ti object spectroscopy with FORS at the ESOVery Large Telescope, in: Instrumentationin Astronomy VIII, eds. C. Crawford and E.R.Craine, SPIE Proceedings Vol. 2198, p. 317,1994.

3. Böhnhardt, H., Möhler, S., Hess, H.-J.,Kiesewetter, S., Nicklas, H.: DesignBenchmarks of the FORS Instrument for theESO VLT, in: Scientific and EngineeringFrontiers of 8-10m Telescopes, eds. M. Iyeand T. Nishimura, Universal AcademicPress Inc. Tokyo, p. 199, 1995.

4. Möhler S., Seifert, W., Appenzeller, I.,Muschielok, B.: The FORS Instruments forthe ESO VLT, in: Calibrating and Under-standing HST and ESO Instruments, ed. P.Benvenuti, ESO Conference and WorkshopProceedings No. 53, p. 149, 1995.

5. Fürtig, W., Seifert, W.: A set of grisms forFORS, in: Tridimensional Optical Spec-troscopic Methods in Astrophysics, eds. G.Comte and M. Mercelin, ASP ConferenceSeries Vol. 71, p. 27, 1996.

6. Appenzeller, I., Stahl, O., Kiesewetter, S.,Kudritzki, R.-P., Nicklas, H., Rupprecht,G.: Spectroscopy of faint distant objects withFORS, in: The Early Universe with theVLT, ed. J. Bergeron, ESO AstrophysicsSymposia Proceedings, Springer-Verlag, p.35, 1997.

7. Szeifert, T., Appenzeller, I., Fürtig, W.,Seifert, W., Stahl, O., Böhnhardt, H.,Gässler, W., Häfner R., Hess, H.-J., MantelK.-H., Meisl, W., Muschielok, B., Tarantik,K., Harke, R., Jürgens, P., Nicklas, H.,Rupprecht, G.: Testing FORS – the firstFocal Reducer for the ESO VLT, in: OpticalAstronomical Instrumentation, SPIE Proc.Vol. 3355, p. 20, 1998.

8. Nicklas, H., Böhnhardt, H., Fürtig, W.,Harke, R., Hess, H.-J., Jürgens, P.,Muschielok, B., Seifert, W., Stahl, O.,Tarantik, K.: Image motion and flexurecompensation of the FORS Spectrographs,in: Optical Astronomical Instrumentation,SPIE Proc. Vol. 3355, p. 93, 1998.

6

Figure 9: Circular spectro-polarimetry of the magnetic white dwarf GD229. The amount of cir-cular polarisation is shown on the top and compared to the flux spectrum below.

I. Appenzelleriappenze@mail.lsw.uni-heidelberg.de

7

ISAAC, the cryogenic infrared spec-trometer and array camera developed bythe Instrumentation Division at ESO(http://www.eso.org/instruments/isaac/)saw first light on the VLT on the night of16/17 November 1998. Figure 1 showsthe instrument mounted on the NasmythB focus of Unit Telescope 1. Visible arethe 1.6-m diameter vacuum vessel sur-rounding the 350-kg cooled optical as-sembly and the co-rotator system whichcarries the electrical cables and closed-cycle cooler hoses to the instrument androtates with the telescope adapter as thetelescope tracks objects on the sky.ISAAC has 0.9–2.5 µm (SW) and 2–5 µm(LW) ‘arms’ providing imaging and spec-troscopic capabilities over a maximumfield of 2.5′ with a variety of pixel scalesand resolving powers.

The moment of first light was preced-ed by intense activity by the ISAAC teamfrom Garching, assisted by their Paranalcolleagues, to install this complex in-

ISAAC Sees First Light at the VLTA. MOORWOOD, J.-G. CUBY, P. BIEREICHEL, J. BRYNNEL, B. DELABRE, N. DEVILLARD,A. VAN DIJSSELDONK, G. FINGER, H. GEMPERLEIN, R. GILMOZZI, T. HERLIN, G. HUSTER, J. KNUDSTRUP, C. LIDMAN, J.-L. LIZON, H. MEHRGAN, M. MEYER, G. NICOLINI, M. PETR, J. SPYROMILIO, J. STEGMEIER

ESO, Garching and Paranal

GFigure 1: ISAAC as it is nowmounted on the UT1 Nas-myth B adapter-rotator. The1.6-m-diameter vacuum tankhouses the 350-kg cryogenicoptical assembly which iscooled by means of twoclosed-cycle coolers. Variousgas and liquid hoses plus allthe electrical cables are car-ried by the co-rotator systemvisible on the left which ro-tates with the adapter as ob-jects are tracked on the sky.

FFigure 2: Colour compositeimage of the RCW38 star-forming complex obtainedby combining short Z(0.9µm), H(1.65 µm) andKs(2.16 µm) exposures.Stars which have recentlyformed in clouds of gas anddust in this region about5000 light-years away arestill heavily obscured andcannot be seen at opticalwavelengths but become vis-ible at infrared wavelengthswhere the obscuration issubstantially lower. The dif-fuse radiation is a mixture ofstarlight scattered by thedust and atomic and molec-ular hydrogen line emission.The field of view is 2.5 × 2.5′with North at the top andEast to the left and the imagequality is set by the FWHMseeing of 0.4″.

8

Figure 3: A Ks (2.16 µm) image centred on the distant ra-dio galaxy MRC0316-257 (z = 3.14) obtained primarily tolocate other distant galaxies for future spectroscopic ob-servations with ISAAC. This image was obtained by com-bining 45 1-min exposures, taken with the telescope ran-domly offset by small amounts in between ("jittering") toallow subtraction of the bright sky emission. The field mea-sures 2.5 × 2.5 arcmin with North at the top and East tothe left. The seeing was 0.5 arcsec and the limiting mag-nitude is Ks p 22. E

Figure 4: Colour composite image of the galaxy clusterCL2244-02 (z = 0.33) obtained by combining a 20-min jit-tered ISAAC Ks (2.16 µm) exposure with 15-min V and Rexposures made with the VLT Test Camera at the UT1Nasmyth focus. In addition to the prominent blue arc, pro-duced by gravitational lensing of a galaxy at redshift z =2.24, there are also notable, very red arcs, both closer tothe centre and further out which were only detected in theinfrared image indicating that these lensed galaxies areeither not forming stars or are much more distant. The fieldsize is about 1.5 × 1.5 arcmin with North at the top left andEast at the lower left corner. The average seeing wasaround 0.5 arcsec. H

strument on the telescope and start theevacuation and cool-down procedure.Despite its large mass, cooling is actual-ly achieved in only 24 hours by using aliquid-nitrogen pre-cooling circuit in ad-dition to the closed-cycle coolers. Onceoperational and after focussing and align-ing the instrument and telescope pupils,short exposures in the 1–2.5 µm filtersquickly demonstrated the excellent com-bination of Paranal seeing and telescopeplus instrument optical quality by deliv-ering images with around 0.3″ FWHMacross the 2.5′ field. During the next sev-eral nights, test images and spectra of avariety of astronomical targets were madeto exercise and establish the perfor-mance of the various SW instrumentmodes. Due to a few technical problemsencountered in this period, it was thennecessary to dismount and open the in-strument which, although unfortunate,was a possibility for which contingencyhad already been foreseen in the plan-ning of this first test. Four days later, test-ing of the long-wavelength arm startedand is continuing as this short report isbeing written in the control room onParanal.

A first impression of the capabilities andperformance offered by this new instru-ment is provided by Figures 2, 3, 4 and5, which show a small selection of the im-ages obtained so far. More images plusthe first spectra obtained can be viewedat http://www.eso.org/outreach/press-rel/pr-1998/pr-19-98.html

9

Figure 5: Combination of 15-min Ks (2.16 µm) and L(3.8 µm) band images of the central regionof the galaxy NGC1365 made with the Long Wavelength arm plus chopping at the telescopesecondary mirror. The field is 16 × 16″ with North at the top and East to the left and shows theprominent Seyfert nucleus plus a rich complex of star forming regions extending over a regionabout 1 kpc across and including one almost at the nucleus to the East.

News from VLT Science VerificationThe data from the UT1 Science

Verifications (SV) have been releasedto the ESO community on October 2,1998, with the data relative to the Hub-ble Deep Field South being availableworldwide. The SV data can be retrievedeither from the ESO Web server athttp://www.eso.org/paranal/sv or ask thelibrary of your institution for the CD-ROMset containing the combined and the rawdata. These CD-ROM sets where mailedin October to all astronomical instituteswithin ESO member states (addressed tothe institute’s library). Since some sets arestill available, interested people may getone by contacting ESO through the aboveWeb page.

A widely publicised editorial of theBoard of Directors of Astronomy andAstrophysics has called for papers basedon the UT1 SV data, with those submit-ted by November 30, 1998, and passingthe peer referee process, being publishedin a special issue of A&A Letters on March1, 1999.

With the successful installation andcommissioning of FORS-1 at UT1 and the

forthcoming commissioning of ISAAC,the plans for the SV observations withthese instruments are in an advancedstage of preparation. The FORS-1 SVTeam has been formed (based large-ly on a new set of astronomers com-pared to UT1 SV Team), and the ISAACSV Team is being assembled. The targetselection for SV will reflect the expecta-tions of the community, with an attemptbeing made to cover as far as possiblethe large variety of astronomical ar-eas in which ESO astronomers are cur-rently active.

FORS-1 SV observations will takeplace during the dark time in January(14–21 January, 1999), while ISAAC SVobservations are planned for 18–25February, 1999. The astronomical targetsof SV will be advertised through theabove SV Web in advance of the obser-vations. The two SV data sets should bereleased about one month after theywere obtained, as it was done for the UT1SV data obtained with the Test Camera.The SV Teams will make every effort toprovide combined and flatfielded data.

Along with the SV data, also sciencegrade data obtained during instrumentcommissioning may be released. Keepwatching the Science Verification Webpages for any news.

The instrument SV observations will in-clude some of the observations for whichthe VLT was specifically built. With FORS,the optical spectroscopy of faint objects– especially in multi-object mode – as wellas the deep imaging on a respectablefield of view will demonstrate the capa-bilities of the VLT to a much larger degreethan could be achieved by the TestCamera images.

With ISAAC, the start of near-infraredobservations with a large telescope inmany ways will open up completely newcapabilities. By having prompt access toFORS and ISAAC science grade data as-tronomers in the community will haveyet another, even more exciting opportu-nity of using and familiarising with VLTdata, specifically of the kind that will beprovided by UT1 to the community of suc-cessful investigators in ESO observingPeriod 63.

Performance M2 Mirror #1 M2 Mirror #2

Radius of curvature 4554.62 mm 4554.91 mmConic constant –1.66980 –1.67046

Optical Quality in Passive mode 0.32 arcsec RMS slope error 0.20 arcsec RMS slope errorOptical Quality in Active mode 18.0 nm RMS (Wavefront) 15.7 nm RMS (Wavefront)Central Intensity Ratio (min.) CIR = 97.4% CIR = 98.1%Microroughness 1.5 nm RMS 1.8 nm

During the month of September 1998REOSC Optique in Paris successfullycompleted the mechanical mounting andthe optical tests of the second VLT M2mirror in Beryllium. The secondary mir-ror was accepted by ESO and shipped toDornier Satellitensysteme for integrationin the second Electromechanical Unit,undergoing its final dynamic acceptancetesting.

The secondaries of the VLT, with theirdiameter of 1116 mm, are amongst thelargest Beryllium mirrors ever produced.The manufacturing process, set up forthe first mirror by REOSC and its sub-contractor Brush Wellmann, is optimisedto achieve the low mass and inertia de-manded by the steering capability ofthe mirrors, without compromising theoptical quality and the long-term opticalstability. The manufacturing process, al-ready described in a previous issue of TheMessenger (No. 86 of December 1996),has been further optimised after theproduction of the first mirror. The blanksare made out of a Beryllium billet ob-tained by Hot Isostatic Pressing of struc-tural grade Beryllium powder (I-220H)lightweighted by machining pockets in theback surface. Due to the hardness ofthe material and the depth of the pock-ets, there is the risk of vibration of the tool,possibly leading to its rupture. Therefore,

the material removal is done at veryslow rate.

The total machining time of the backand front faces – the latter up to less than20 µm tolerance from the final asphere –takes approximately 9 months. After thegeneration of the asphere by grind-ing, the blank is Nickel plated on bothsides. The polishing is done on the Nickellayer. As the secondary mirror of the VLTis undersized, the optical surface extendsup to the mechanical edge. To polish upto the edge, a lip was left at the outer rimof the first mirror, that was later cut byelectromachining discharge (EDM).Although the internal stresses in the mir-ror were very low (estimated around 5MPa), their release, after cutting off thelip, generated a warping which demand-ed further polishing.

For the second mirror REOSC has de-veloped a method to polish up to the bor-der without producing a turned downedge. The lip was therefore cut beforeNickel plating and the subsequent pol-ishing.

Some of the most relevant physicaldata of the mirrors are provided in Table 1.

The lower mass of the second mirroris due to a stricter control of the di-mensional tolerances during machiningand to a slight difference in the values ofthickness of the Nickel plating. The vari-

ation of the mass and inertia are wellwithin the tuning capability of the controlsystem for the steering mechanism of thesecondary unit, so that mechanical ad-justment is not needed. The third andfourth mirrors (presently in production) willhave characteristics similar to the secondone.

The mirrors have been tested after in-tegration in their cells and with the opti-cal face down, therefore under conditionssimilar to those encountered in the tele-scope. The mirrors are tested against aZerodur matrix. The measurements areperformed with the mirror mounted in itscell, and with the cell fixed onto a templatesimulating the eletromechanical unit. Theradius of curvature and the conic constantof the matrix were previously measuredwith a Hartmann test using a null-lens,and then with a second independentnull-lens for cross-check.

The specification demanded by ESOcontains requirements for the so-calledpassive and active modes. In the passivemode the full pupil of the mirror is testedand the optical quality is deduced after re-moval of the effects of mirror curvature(focus error) and conic constant (3rd or-der spherical aberration). The slope er-rors are obtained after measurementagainst the matrix.

The active mode quality is expressedin Central Intensity Ratio (CIR) whichcompares the performance of the tele-scope to that of a diffraction-limited tele-scope in well-defined atmospheric con-ditions. The CIR is evaluated after re-moval of the 16 first natural modesof the primary mirror, which are correct-ed by the active optics of the VLT. Theeffects of high spatial frequency errorsare calculated with a sub-pupil test (maskwith sub-apertures on a mirror diame-ter), and their effect subtracted from theCIR to obtain the final minimum value ofthe CIR.

Classical checks such as micro-roughness, cleanliness and other in-

10

T E L E S C O P E S A N D I N S T R U M E N T A T I O N

Performance of the First Two Beryllium SecondaryMirrors of the VLTS. STANGHELLINI, A. MICHEL, ESO

Characteristics M2 Mirror #1 M2 Mirror #2

External diameter 1116 mmDiameter of centre hole 46 mmHeight at mirror centre 170 mmDimension of bevel at edge < 0.2 mm Mirror mass 43.1 kg 41.8 kgTotal mass (M2 + mounts +cell) 52.5 kg 51.2 kgTotal moment of inertia 4.17 kg m2 4.07 kg m2

First eigenfrequency on three rigid points > 635 HzFirst eigenfrequency on the three mounts > 365 Hz

Table 1. Mechanical characteristics of the first two VLT secondaries

Table 2. Optical performance of the first two VLT secondaries

spections were also performed. The sur-face quality of both mirrors is excellentand within specification.

Table 2 provides the major opticalperformance of the first two secondarymirrors. The two mirrors are extremelysimilar, with only a fraction of the budgetof error allocated for the radius of cur-vature and the conic constant being used.REOSC was also able to improve the op-

tical quality with the polishing of the sec-ond mirror.

It is possible to conclude that whilethe feasibility of the Beryllium technolo-gy used had been demonstrated withthe delivery of the first secondary mirrorand the excellent images already obtainedby the first VLT telescope, a further ad-vance in the manufacturing processwas achieved by REOSC with the recently

delivered second mirror, leading to animprovement of the optical performanceand to a shorter production time. The to-tal production of the first mirror demand-ed more than two years, while the sec-ond mirror was produced in around 20months.

11

sstanghe@eso.org

During the last three months, the op-eration of the NTT has been particularlysmooth; we did not experience any ma-jor problem, and the weather has beenfairly cooperative. The technical down-time was of the order of 2%.

On August 12 and 13, the NTT and itsteam passed a very detailed “AcceptanceReview”, during which all the technicaland operational aspects of the telescopewere presented. This review constitutedthe formal return of the NTT from the VLTDivision to the La Silla Observatory after

the Big-Bang. The review board found thetechnical and operational status of thetelescope to be more than satisfactory;the staff, system, operation, and proce-dures all received excellent reviews.

The dewar of SUSI2 was suffering fromvacuum losses since the installationof the instrument (cf. last issue of TheMessenger). It has been exchanged bya new, improved dewar, which immedi-ately worked perfectly. Also, we identifiedthe source of a mysterious light contam-ination which occasionally affected the im-

ages: the lamps of the rotator and altitudeencoders were not perfectly shielded. Insome positions of the rotator, a part of theinstrument which was not correctly black-ened was reflecting the encoder light tothe detector. Additional light baffles havebeen mounted, and the blackening of theinstrument completed, resulting in thecomplete removal of the contamination.SUSI2 is now performing at its expectedlevel.

Over the past couple of months, theObservation Block (OB) database in La

The La Silla News Page

News from the NTTO. R. HAINAUT

The editors of the La Silla News Page would like to welcome readers of the twelfth edition of a page devoted toreporting on technical updates and observational achievements at La Silla. We would like this page to informthe astronomical community of changes made to telescopes, instruments, operations, and of instrumental perform-ances that cannot be reported conveniently elsewhere. Contributions and inquiries to this page from the commu-nity are most welcome. (J. Brewer, O. Hainaut, M. Kürster)

Silla, the OB repository in Garching, andthe communication between these twohave caused many problems, some-times resulting in a degradation of theoperation. While these problems wereefficiently tackled by the Data Manage-ment Division, they nevertheless re-vealed some weaknesses in the system,which have already been partially solvedby improving the structure of the OB data-base. A more robust version of the P2PPsoftware (used at the telescope for thepreparation of the OBs) will soon be de-livered.

Since the “Big-Bang”, the operation ofthe NTT is performed according to de-tailed and optimised procedures andcheck-lists; these have helped ensure thehigh reliability of the system. While these

procedures are in constant evolution,the vast majority of the NTT operation isnow documented as an extensive col-lection of WWW-based procedures. Thenext step is to tackle the maintenanceplan of the system; while most of themaintenance tasks are regularly per-formed, their execution and the corre-sponding procedures are currently most-ly left to the experience of the team’s tech-nicians (fortunately, their experience is ex-tensive). The maintenance plan and itscorresponding procedures are now beingdocumented in the same way as wasdone for the operation procedures. In ad-dition, a vast collection of “templates”(scripts that can be automatically exe-cuted by the whole system) has been de-veloped to measure the various para-

meters needed for the maintenance,such as the current consumed by thevarious motors of the rotators, the in-strumental flexures and focus variations,etc. A first calibration plan will soon be im-plemented, and will continue to be de-veloped and expanded in 1999. This willallow us to detect problems at an earlystage, and take corrective actions beforethey lead to a loss of observing time.

Finally, let me announce that new ver-sions of the EMMI and SOFI user’s man-uals have been released and are avail-able on the NTT web page. They are verydetailed, and include recent throughputmeasurements of all observing modes, aswell as notes for the preparation of the ob-servations with the “neoclassic” system(Observation Blocks, etc.).

12

2.2-m Telescope Upgrade: a Status ReportT. AUGUSTEIJN

The hardware modifications andthe installation of the VME-based tele-scope control system (TCS) were fin-ished according to schedule on the1st of October 1998. In addition, thetelescope cabling has been cleaned, thecontrol room refurbished, and the TCSand the instrument control system(DAISY+) workstations have been in-stalled and integrated into the local net-work.

At the time of writing, a direct camerais mounted at the Cassegrain focus fortest purposes. We are currently in theprocess of testing the TCS user interfaceand verifying the telescope performancein preparation for the arrival of the WideField Imager (see previous issue of TheMessenger). Although the current set-upis not optimal, we regularly obtain sub-arc-second seeing and on various occasionshave obtained seeing below 0.5 arcsec-

onds, demonstrating the excellent opticalquality of the telescope. In combinationwith the well-sampled half-degree field ofthe WFI, the 2.2-m Telescope will be amajor asset to the observing facilities atESO. Updated information on the 2.2-mTelescope and the WFI is available on the2p2 Team Web Page (see URL:http://www.ls.eso.org/lasilla/Telescopes/2p2T/2p2T.html).

Performance of CES 3.6-m Fibre Link and Image SlicersM. KÜRSTER

Three new image slicers for the fibrelink from the Coudé Echelle Spectrometer(CES) to the Cassegrain focus of the 3.6-m telescope have arrived. Their proper-ties are summarised in the following tablewhich compares the measured andplanned resolving powers, and lists thenumber of slices produced by each sliceras well as the extent on the CCD of thetotal spectrum (all slices) in the directionperpendicular to the dispersion.

Even though the goal for the maximumresolving power has been missed by12%, the high-resolution image slicerdoes make the CES by far the highest-res-

olution facility among all ESO instru-ments. Using the spectrograph entranceslit instead of an image slicer even high-er resolving powers up to R = 284,000 canbe achieved, albeit at the expense of alarge loss in throughput.

During the commissioning period (3–7November 1998) strong and variable cir-rus clouds hampered the determinationof reliable measurements of the efficien-cy of the whole optical train (telescope,fibre link, image slicers, spectrograph anddetector). Nevertheless, it was possibleto obtain some encouraging lower limitsfor a few wavelengths that were not

much lower than the expected values. Asan example, the efficiency near 6100Å islisted in the table for CCD #38 and lessefficient CCD #34 (values in brackets).Within a few percent, all slicers yield thesame throughput. Results from final effi-ciency measurements will be communi-cated when they are available.

The CES instrument is currently un-dergoing a refurbishment which we ex-pect to be finalised by the end of July1999. The major modifications foreseenfor the CES include:

• An upgrade of the instrument controlsystem to VLT standards (second VLTcompliant 3.6-m instrument afterEFOSC2).

• Stabilisation of the predisperser viaa new drive system.

• Improving the turntable drive.• Stabilisation of the thermal environ-

ment of the instrument by modifying thecoudé room.

We will continue to report all importantchanges on these pages.

Slicer Resolving power # Extent Efficiency at 6100 ÅAchieved Planned Slices (pixels) Lower limit Expected

High res. 194,000 220,000 12 400 7.2% (3.6%) 8% (4%)Medium res. 138,000 110,000 8 265 " "Low res. 88,000 80,000 6 168 " "

13

Abstract

The Large Southern Array is one of thehighest priority projects in European as-tronomy today. The planned submillime-tre operating wavelengths and the needto maximise the efficiency of observingtime make the selection of the observa-tory site a fundamental issue.

1. Introduction

The last decades have seen greatprogress in (sub-)millimetre astronomy,leading to the design and construction ofnew and powerful instruments. There arenow many very good 10–15-m-diameter(sub-)millimetre telescopes available tothe astronomical community. In particu-lar, the 15-m-diameter Swedish-ESOSubmillimetre Telescope (SEST) isunique in the Southern Hemisphere. Thistelescope has opened the southern skiesto millimetre astronomy, achieving out-standing scientific results during the lastten years (see articles in The MessengerNo. 91, 1998).

The continuing quest for higher reso-lution and greater sensitivity calls for ra-

dio telescope arrays. In 1995, Europeaninterest in a millimetre array sparked theLarge Southern Array project (LSA) withan agreement between ESO, the Institutde Radio Astronomie Millimétrique(IRAM), the Onsala Space Observatory(OSO), and the Netherlands Foundationfor Research in Astronomy (NFRA). Withthe aim of achieving a total collecting areaof 10,000 m2 and an angular resolutionbetter than 0.1 arcsec at a wavelength of3 mm, this project was at the time themost ambitious plan considered for mil-limetre astronomy. Since the early 1980s,the US National Radio AstronomyObservatory (NRAO) has planned an-other array, the Millimeter Array (MMA)which is a 2000-m2 array capable of work-ing at sub-millimetre wavelengths. Bothprojected instruments would be sited innorthern Chile, naturally suggesting col-laboration.

In 1997, ESO and NRAO agreed towork together to explore a common mil-limetre/submillimetre array, to be financedjointly by Europe and the US. TheLSA/MMA project is a compromise arraywith 64 telescopes, each of 12 m diam-eter, giving a total collecting area of

7,000 m2, with baselines up to 10 km, andwith sub-millimetre capabilities. It will belocated on Llano de Chajnantor, a 5000-m-high plateau near the village of SanPedro de Atacama in the Atacama Desertarea. This will be the most powerful(sub-)millimetre telescope in the world,opening completely new horizons in as-tronomy.

The LSA/MMA will be the natural com-plement, in its wavelength range, to theHST and VLT, with equivalent resolutionand sensitivity and the additional capa-bility to observe the inner part of the dust-obscured regions of star formation anddust enshrouded galaxies.

At (sub-)millimetre wavelengths, the at-mosphere presents natural limits to thesensitivity and resolution of astronomicalobservations. Pressure-broadened mol-ecular lines, in particular of water vapour,are the cause. The atmosphere adverselyaffects observations both by attenuatingthe incoming signal and by increasing thesystem noise, since it radiates thermal-ly. Furthermore, inhomogeneities in thewater vapour distribution change theelectrical path length through the atmos-phere. These random variations result in

N E W S F R O M T H E L S A

European Site Testing at Chajnantor: a StepTowards the Large Southern ArrayA. OTÁROLA1, G. DELGADO 2, R. BOOTH 3, V. BELITSKY 3, D. URBAIN 3, S. RADFORD 4, D. HOFSTADT 1, L. NYMAN 2, P. SHAVER 1, R. HILLS 5

1European Southern Observatory; 2SEST/Onsala Space Observatory; 3Onsala Space Observatory; 4National Radio Astronomy Observatory; 5Mullard Radio Astronomy Observatory

Figure 1: Partial view of the site-testing equipment deployed in Llano de Chajnantor. The LSA container is to the left, while the MMA containeris to the right. In the background Cerro Chascón.

14

phase errors between the array ele-ments which degrade the sensitivity andresolution of the images. To minimisesuch atmospheric degradation, then, theideal site should be a dry plateau at highaltitude.

For several years the European col-laboration has been exploring the north-ern area of Chile searching for potentialsites (Otárola et al., 1996). The experi-ence has helped not only our own efforts,but also other groups conducting sitestudies and/or astronomical projects in thesame area. NRAO began site character-isation measurements at Llano deChajnantor in 1995. In 1998, after morethan a year of measurements at anothercandidate site (Pampa Pajonales), theEuropean measurements started atChajnantor as well. Several instrumentsare deployed alongside those of NRAO(Fig. 1). The goal is to share the infra-structure and to correlate the data ob-tained at the same place and time withthe different instruments.

The work done at Chajnantor by NRAOshows that the median zenith opticaldepth at a frequency of 225 GHz isabout 0.06 (see Fig. 2). This is significantlybetter than that measured at the CSO onMauna Kea with an identical instrument.Also, there is less diurnal variation intransparency on Chajnantor than onMauna Kea. These data span the period1995 to 1998. The main meteorologicalevents (“Altiplanic Winter” and “El Niño –ENSO) have had an adverse effect on the

data, as can be seen in Figure 2.However, the Chajnantor site provides ex-cellent observing conditions, perhapsamongst the very best in the world for sub-millimetre wavelengths. The high atmo-speric transparency allows high-efficien-cy telescope operations.

Moreover, the whole area aroundCerro Chascón, including Llano deChajnantor, has recently been declaredto be of “scientific interest” by a presi-dential decree. A land concession hasbeen granted to the Comisión Nacionalde Investigación Científica y Tecnológica(CONICYT), the agency in charge ofevaluating, supporting and funding sci-entific and technological research inChile. The area around the LSA/MMA is

Figure 2: Quartiles of the zenith optical depth– opacity – at 225 GHz (upper plot) and thephase stability (lower plot) observed by NRAOduring the years 1995–1998. The globalweather conditions Altiplanic Winter and oneEl Niño – Southern Oscillation (ENSO) eventcan be seen in the cumulated data.

Figure 3: Map of the Atacama Desert regionin Chile (scale 1:3,000,000). To the east of“Salar de Atacama” can be seen the locationof Llano de Chajnantor, while to the south ofAntofagasta Paranal can be seen (adaptedfrom “Atlas de la República de Chile”, InstitutoGeográfico Militar, 1983).

15

now protected from third parties andfrom possible mining claims or othersuch activities.

In this report an overview of the site,the instruments used to characterise it,their scientific rationale and some pre-liminary results are presented.

2. Chajnantor

Llano de Chajnantor is the selected sitefor the combined LSA/MMA project. Theatmosphere at Chajnantor and the near-by plateau “Pampa La Bola” have beenstudied extensively by NRAO and theNobeyama Radio Observatory (NRO), re-spectively. Measurements have shownthat this area is among the driest in theworld, with conditions that excel for as-tronomical research at several wave-lengths.

Chajnantor is a 5000-m-high plateauabout 60 km ESE of San Pedro deAtacama. This village is a place of greatethnological and archaeological impor-tance, being the oldest continuously in-habited site in Chile. It is also one of themost important tourist attractions of theregion because of the impressive moun-tain and desert landscapes, the pre-Columbian archaeological sites, the finearchaeological museum, and the relaxingatmosphere given by the rustic nature ofits centuries old adobe houses and nar-row streets.

2.1 Geomorphology

Three north-south mountain rangesdominate the geomorphology of theAtacama Desert area (see Fig. 3): theCoastal mountains, with peaks over3500 m, the Domeyko mountains, locat-ed at middle longitudes, and the Andes,with peaks well above 5000 m.

The 25 km2 Chajnantor plateau is lo-cated on the western edge of the Andesat 67°45′ west and 23°01′ south, about380 km NE of Paranal. This plateau isopen to the west and north-west, but sev-eral 5600-m peaks border the site in oth-er directions.

2.2 Logistics

Prom Pedro de Atacama an interna-tional highway to Argentina, “Paso deJama”, allows an easy access to thesite. By March 1999 this road will bepaved through the CONICYT land con-cession, ensuring access by a goodpaved road.

A power source is essential for the de-velopment of the LSA/MMA. The nearestexisting power lines are 180 km NW inCalama. Due to the growing commercialrelations with Argentina, however, severalprojects are being carried out to bring nat-ural gas from Argentina to produce elec-trical power for the Atacama region. Oneof these gas pipelines crosses the Chaj-nantor plateau. NRAO and GasAtacama,the pipeline owner, have signed an agree-ment to install a gas tap at the north edgeof the Chajnantor area. This will allow the

LSA/MMA to generate electrical power asneeded.

A few years ago, when we first start-ed exploring this area, Chajnantor was avery isolated place. By early 1999, the sitewill have a good paved road running near-by and natural gas on site.

2.3 Climate

Weather conditions in the AtacamaDesert are dominated by a high-pres-sure anticyclone, an inversion layeron the coast, the north-south coastalmountains, the Andes mountains, the high

insolation, and the high ventilation.The whole area is extremely dry and

lies in the meridional transition region. Tothe north, the Altiplano convective sum-mer rains predominate, where frontalwinter rains prevail to the south. The arid-ity originates in the prevailing anticycloneconditions associated with a slow down-ward air motion (Fuenzalida et al. 1996).Over the coast this subsidence is re-sponsible for an inversion layer severalhundred metres thick that starts about1000 m. The inversion layer separates themoist air of the lower marine boundary lay-er from the dry subsiding air and acts as

Figure 5: Diurnal cycle of the solar radiation for the period June–September 1998 in the Chajnantorarea.

Figure 4: One of the 183-GHz water vapour monitors to the right of the antennas of the 11.2 -GHz interferometers.

16

a barrier to any mixing process thatmight transport moisture evaporatingfrom the sea to upper levels. The well-mixed marine boundary layer is restrict-ed to lowlands along the coast onthe west slope of the coastal mountainrange.

As humid air masses moving westwardfrom the Atlantic zone rise up the eastslope of the Andes they cool and precip-itate. Hence, very little humid air gets overto the west side of the Andes. The ex-ception is during summer, when the high-pressure system over the Pacific weak-ens and, at the same time, moves south-ward (Fuenzalida et al. 1987, Aceitunoet al. 1996). This change in atmospher-ic circulation allows warm humid air fromthe Amazon to reach the Altiplano innorthern Chile and produce rain. This isknown locally as “Bolivian or AltiplanicWinter”.

The driving force behind the climato-logical processes is the extremely high so-lar radiation. The daily and yearly maxi-ma, averages and ranges of the insola-tion are amongst the highest recorded inthe world. The average daily maximumreaches almost 1300 W/m2 during sum-mer. Only convective and advective

processes can remove the absorbedheat because the extreme aridity meansthat the evaporating water cannot dissi-pate the heat. The wind then is the mainenergy carrier. The strong thermal con-trast between the Pacific and the Andesdrives a constant sea-mountain airflow.Almost every climatic process within theboundary layer is controlled by this re-gional wind system. At night the bound-ary layer regularly disappears so the cli-mate near the ground is directly influencedby the upper atmosphere winds (Schmidt1997).

3. Site Testing Equipment

At the beginning of June 1998, theOnsala-ESO European site-testing groupdeployed equipment to characterise theatmospheric conditions at Chajnantor for(sub-)millimetre astronomy. Our instru-mentation and measurements comple-ment the measurements carried out bythe other groups (see Fig. 4).

3.1 Container

The computers and other ancilliaryequipment are installed in a 20-foot in-

sulated ocean-shipping container. About18 m2 of photovoltaic panels are mount-ed outside to charge a battery bank thatprovides about 700 W around-the-clock.The computers housed in the containercontrol the instruments and store the data.Communications are maintained with ananalogue cellular telephone. Data for allthe instruments are retrieved about oncea month during routine visits.

3.2 Weather stations

Two weather stations have been in-stalled on the plateau. They measure airtemperature, solar radiation, atmospher-ic pressure, relative humidity, and windspeed and direction. Their main charac-teristics are detailed in Table 1.

The weather stations include a datalogger that, with the actual samplingrate, can store 42 days of data. The windsensors are atop a 4-m mast and the oth-er instruments are installed in a ventilat-ed shelter to protect them from direct so-lar radiation.

Analysis of these data improves our un-derstanding of the climatology of thesite, with emphasis on the wind speed anddirection and the diurnal and seasonalpatterns. These data also provide infor-mation on the environmental constraintswhich will influence the design of theLSA/MMA antennas, and it is a basic cal-ibration input for the analysis of the dataobtained with other instruments at the site.

Figure 6: Cumulative function and histogram for the wind speed over the period June–September1998 in the Chajnantor area. Upper plot shows night-time while lower one is day-time.

Figure 7: Wind speed and direction over 14hours during the day (upper plot), and his-togram with the percentage of time that thewind blows in a particular direction during thissame time period (lower plot). The distributionfor night-time is similar.

Sensor Operating range Accuracy

Air temperature –30 – +70 °C 0.2 °CRelative humidity 5–95% 1.5% in the range 5% to 60% Solar radiation 0–1100 W/m2 ±5% per MJoule, spectral response 400–1100 nmAnemometer 0–150 km/h 0.5 m/sWind direction 0–360 degrees 3.0 degreesBarometric pressure 740–490 mbar ±5 mbar

Table 1: Main characteristics of the sensors available in the weather stations deployed inChajnantor

17

3.3 Interferometer

Atmospheric phase stability, or “seeing”,is a crucial issue for the LSA/MMA.Because radio waves travel more slow-ly in wet air than in dry air, fluctuations inthe water vapour distribution will causevariations in the electrical path lengththrough the atmosphere. Path lengthvariations across the array aperture willdegrade both image quality and array sen-sitivity. These path length fluctuations,which are almost independent of ob-serving frequency at millimetre wave-lengths, correspond to phase fluctuationsthat scale linearly with frequency. Simu-lations suggest that phase fluctuationsless than 10 degrees rms at the observ-ing wavelength will have little impactupon most images, phase errors of 30 de-grees rms will permit imaging with up to200:1 dynamic range, albeit with some-what reduced sensitivity, and image re-construction becomes all but impossiblewhen phase errors exceed 60 degreesrms (Holdaway and Owen 1995b). Al-though the LSA/MMA will undoubtedlyuse a correction technique to overcomeatmospheric image degradation (eitherfast switching or radiometric phase cali-bration), the better the natural stability, thebetter these correction schemes will work(Holdaway et al. 1995c, Woody et al.1995).

Atmospheric phase stability at Chaj-nantor is monitored with two 300-m base-line interferometers observing 11.2 GHztracking beacons broadcast by geosta-tionary communications satellites (Rad-ford et al. 1996). One instrument was in-stalled by the MMA project in May 1995and another was installed by the LSA pro-ject at Pajonales in April 1997 and latermoved to Chajnantor in June 1998.Because the atmosphere is non-disper-sive away from line centres, these mea-surements can be extrapolated to char-acterise the atmospheric phase stabilityup to at least 350 GHz and, with care,throughout the submillimetre region.These interferometers sense atmosphericstructures on 300-m and smaller scales.The phase stability is characterised by therms phase fluctuations calculated over 10-minute intervals. This interval is twentytimes longer than the time it takes an at-mospheric feature to move the length ofthe baseline at 10 ms–1, which is the me-dian wind speed aloft. Thermal instru-mental phase noise is on the order of 0.1degree rms at 11.2 GHz, while the small-est observed atmospheric phase fluctu-ations are – after correcting for instru-mental noise – 0.3 degree rms (Holdawayet al. 1995c).

The two interferometers are collo-cated along the same east-west 300-mbaseline, but observe different satellitesseparated by about 5°. As the (quasi-static) pattern of atmospheric turbu-lence passes over the interferometers,they will record similar signals that aredelayed from one to the other. From thisdelay, some estimate of the wind speedaloft, and the geometry, we can deter-

mine the effective height of the turbu-lent layer. Differences in the interfer-ometer signals, on time scale shorterthan the delay, provides a measure ofthe evolution of the turbulence (Hold-away and Radford 1998). The effec-tiveness of fast switching at overcom-ing atmospheric image degradation de-pends on both the height and the timescale for evolution of the turbulence(Holdaway and Owen 1995b).

3.4 183 GHz water vapour monitor

Now under development at several ob-servatories, water-line radiometry is apromising technique for overcoming theeffects of atmospheric phase fluctua-tions. The varying depth of the watervapour column is determined by contin-uously monitoring the intensity and shapeof a water-vapour emission line. Changesin the atmospheric phase delay are then

Figure 8: Diurnal temperature cycle over the period June–September 1998 in the Chajnantorarea.

Figure 9: Cumulative function and histogram for the relative humidity over the periodJune–September 1998 in the Chajnantor area.

18

determined and used to improve theastronomical images. Both the 22 GHzand the 183 GHz line are candidates forthis method, but it is not yet clear whichis best for the conditions at Chajnantor.

Two radiometers for the 183 GHz wa-ter-vapour line have been built as a col-laborative project between Onsala SpaceObservatory, through the Group forAdvanced Receiver Development atChalmers University of Technology, andthe Mullard Radio Astronomy Observatory(MRAO), Cavendish Laboratory, Cam-bridge. They have been installed at Llanode Chajnantor at the ends of the 300-mbaseline of the 11.2 GHz interferometer.The plan is to compare the atmosphericphase fluctuations determined from theradiometers and from the interferometeras a test of the phase correction scheme.The radiometers also allow, under com-puter control, complete movement of thebeam across the sky.

These radiometers are double-sideband heterodyne receivers with three de-tection bands spaced 1.2, 4.5, and 7.5GHz away from the line centre. Thethree-channel spectrum measures thewater line shape and intensity. Theamount of precipitable water vapour isthen determined by fitting to an atmos-pheric model (Wiedner 1998).

The mixers are of the sub-harmoni-cally pumped Shottky types developedat the Rutherford and Appleton Labora-tories, in the UK. This design has twomain advantages: it does not require anexternal bias and we do not need to dou-ble the Gunn oscillator frequency topump the mixer. The resulting RF circuitis simpler and more reliable. In the labo-ratory, the receiver temperature was

measured to be about 1500 K. This hasbeen confirmed during normal operationin the field.

The local oscillator (LO) is a free-running, 91.7-GHz Gunn oscillator. It isnot phase locked because of the broad-ness of the water line and because theline is very symmetric. Hence any in-stability is compensated by the DSBdetection. Also the frequency stabilityof the oscillator is better than 4 MHz/°C,so by regulating the LO temperature towithin 1 °C, we obtain a negligible 0.004%frequency drift.

The radiometers are continuously cal-ibrated by switching the beam betweenthe sky and two reference black bodiesat temperatures of 35 and 100 °C.

4. Results

Daily climatology data have been col-lected for the period June–September1998 winter season in the SouthernHemisphere, and are presented as an av-erage over this period. Figure 5 shows thediurnal cycle of the solar radiation forJune–September 1998. The radiationpeaks at about 1000 W/m2, and the av-erage maximum is around 800 W/m2. Thehistograms in Figure 6 show that the windis a constant in the area, with lowerspeeds during the night (median of about6 m/s) and higher speeds during the day(median of about 10 m/s). This is a pa-rameter that must be kept in mind duringthe design stage of mechanical structuresto be built in the area, especially the an-tennas. Over 80% of the time the wind

Figure 10: Amount of precipitable water vapour content during the month of September 1998in the Chajnantor area (upper plot), and cumulative function and histogram for the same month(lower plot). Several storms affecting the area can be seen.

Figure 11: Data from the LSA and the MMA 11.2-GHz interferometers (phase stability moni-tors) for August 1998. The good correlation between both instruments shows that they performidentically.

19

blows from the west (from the desert tothe high mountains) as can be seen inFigure 7. The diurnal cycle of tempera-ture is shown in Figure 8: during the night,the temperature becomes stable withthe high solar radiation influencing it dur-ing the day.

Over the observation period the rela-tive humidity did not change much dur-ing a diurnal cycle, being very stable dur-ing the night – at an average of around38%–decreasing during the day to reacha minimum average of around 25%.Figure 9 shows the cumulative functionand histogram for the relative humidity.

Data for the precipitable water-vapour(PWV) content of the atmosphere duringSeptember 1998 are shown in Figure 10,where some days with very bad weath-er can be seen as peaks distribution. Itcan also be seen that normally PWV islower and stable during local night-time.In the same Figure the histogram for PWVshows that 50% of the time PWV is lessthan 1 mm and 75% of the time less than1.7 mm. Since the first analysed monthfor Chajnantor (September) was a par-ticularly bad month, with several stormsover the area, due to a transition betweenthe two global weather conditions knownas “El Niño” and “La Niña”, better resultsare expected after some accumulated

statistics. It should be noted that above3.5 mm of PWV, measurements are nolonger linear due to instrumental limita-tions.

Data from the two interferometers forAugust 1998 show they perform identi-cally. The rms phase fluctuations calcu-lated over 10-minute intervals for each in-terferometer are compared in Fig. 11.Most of the scatter in this figure is due toatmospheric evolution on short timescales. Although the computers control-ling the instruments are synchronised tobetter than 1 s, the 10-minute intervalsused for analysis are only synchronisedto 5 minutes. The MMA interferometermeasures slightly larger fluctuations thanthe LSA instrument, largely because theMMA interferometer observes a satelliteat lower elevation (Holdaway and Ishiguro1995).

A detailed example of the interfer-ometer data (Fig. 12) shows the ex-pected time delay as the (quasi-stat-ic) atmospheric turbulence patternpasses sequentially through the inter-ferometers’ lines of sight. In this exam-ple, the LSA signal leads the MMA sig-nal by about 10 s. Together with thegeometry and an estimate of the windspeed aloft, we can estimate the heightof the dominant turbulent layer. Other

structures are also evident in the data,possibly due to layers at different heightsor to fast evolution of the turbulent pat-tern.

5. Conclusions and Future Work

On Llano de Chajnantor we have in-stalled weather stations, two 183 GHz wa-ter vapour radiometers, and a phase sta-bility monitor (11.2 GHz interferometer).Data are recorded automatically and re-trieved once a month during periodic vis-its to the site. Data processing and analy-sis are in their first stages. The meteo-rological data are analysed in Chile, the183 GHz radiometer data are analysedin Chile and Sweden, and the interfer-ometer data are analysed in collaborationwith NRAO.

The site remains extremely promisingfor submillimetre interferometry althoughbad weather caused by the transition be-tween the two global weather conditionsknown as “El Niño” and “La Niña” was ex-perienced.

In the near future we will compareour data with those obtained throughradio sonde launches – a campaignwhich started in October 1998 and willcontinue through 1999. This is a col-laboration between ESO, NRAO, Cor-

Figure 12: Example of the 11.2-GHz interferometers data showing the expected time delay as the atmospheric turbulence patterns cross overthe interferometer line of sight. From these data a height of about 500 m for the turbulence layer can be inferred.

20

nell University, and the SmithsonianAstrophysical Observatory. As a com-plement to the 183 GHz radiometers, wealso plan to install a 22-GHz radiometerto observe another water vapour lineand correlate measurements. The ra-diometer, on loan from R. Martin fromSteward Observatory, will be installednext year. It has been extensively up-graded to make it more reliable andcapable of automatic operation at thesite.

In the data analysis, we will explore al-ternatives to the atmospheric model fit-ting of the retrieved line from the 183 GHzradiometer measurements, as well as tocorrelate these measurements with oth-er ones done at different frequenciesand/or techniques.

Together with IRAM engineers, we willstart measurements of the wind sampledat a high rate in order to determine thepower spectrum of the wind behaviour.Such information has an important bear-

ing on the design of the LSA/MMA an-tennas.

The data, results and all relevantinformation about the project can nowbe found in the web pagehttp://puppis.ls.eso.org/lsa/lsahome.html

6. Acknowledgements

Many people around the world havemade fundamental contributions to thisproject. We wish particularly to thank theESO staff in Chile, Patricia Adriazola,Viviana Alcayaga, Mary Bauerle, AlfredoCarvajal, Manuel Hervias, Luís Morales,Patricia Parada, and Ivonne Riveros fortheir special assistance whenever wecame with an urgent request.

References

Aceituno P., Montecinos A., Departamento deGeofísica Universidad de Chile, 1996.

The Messenger, No. 91, March 1998.Fuenzalida R., ESO internal report, 1996.

Fuenzalida H., Ruttlant J., in Proceedings ofthe II Congreso Interamericano de Mete-orología, Buenos Aires, Argentina, 1987,6.3.1.

Holdaway M., Ishiguro M., MMA Memo 127,1995.

Holdaway M., Owen F., MMA Memo 136,1995b.

Holdaway M., Radford S., Owen F., Foster S.,MMA Memo 139, 1995c.

Holdaway M., Radford S., MMA Memo 196,1998.

Otárola A., Delgado G., Bååth L., inProceedings of the ESO IRAM NFRAOnsala workshop 11–13/12/95, P. ShaverEd., Springer Verlag, 1996, 358.

Radford S., Reiland G., Shillue B., PASP, No.108, 1996, 441.

Schmidt D., PhD thesis, Friedrich-AlexanderUniversity, 1997.

Wiedner M., PhD thesis, University ofCambridge, 1998.

Woody D., Holdaway M., Lay O., Masson C.,Owen F., Plambeck D. Radford S., SuttonE., MMA Memo144, 1995.

gdelgado@eso.org

The ISAAC Team on Paranal After First Light on UT1

From left to right: J. Stegmeier, G. Fin-ger, A.Moorwood, P.Biereichel, J. Brynell,J.-G. Cuby, J. Knudstrup, M. Meyer, J.-L.Lizon.

21

1. Introduction

To understand the formation of the vis-ible structure in the Universe out of an ini-

tially almost homogeneous matter distri-bution is one of the most fascinatingquests of modern cosmology. The firststep to this understanding is of course

reasons: the present-day structures onthis scale are directly comparable tothe signature of the primordial structuredetected in the microwave backgroundby COBE with similar comoving sizes, andat these scales the observed densityfluctuations are just linear amplifica-tions of the initial conditions in the ear-ly Universe. While the study of the galaxydistribution has already given us veryinteresting insights into the structureon scales of a few hundred Mpc, analternative approach using the next larg-er building blocks of the Universe, galaxyclusters, as probes for the cosmic struc-ture can give us a ready access to even

R E P O R T S F R O M O B S E R V E R S

Probing the Cosmic Large-Scale Structure with theREFLEX Cluster Survey: Profile of an ESO KeyProgrammeH. BÖHRINGER 1, L. GUZZO 2, C.A. COLLINS 3, D.M. NEUMANN 4, S. SCHINDLER 3, P. SCHUECKER 1, R. CRUDDACE 5, S. DEGRANDI 2, G. CHINCARINI 2, A.C. EDGE 6, H.T. MACGILLIVRAY 7, P. SHAVER 8, G. VETTOLANI 9, W. VOGES 1

1Max-Planck-Institut für Extraterrestrische Physik, Garching, Germany; 2Osservatorio Astronomico di Brera, Milano/Merate, Italy; 3Liverpool John-Moores University, Liverpool, U.K.;4CEA Saclay, Service d’Astrophysique, Gif-sur-Yvette, France; 5Naval Research Laboratory, Washington, D.C., USA; 6Durham University, Durham, U.K.; 7Royal Observatory, Edinburgh, U.K.; 8European Southern Observatory, Garching, Germany; 9Istituto di Radioastronomia del CNR, Bologna, Italy

Figure 1: Three of the galaxy clusters not listed in the catalogue by Abellet al. (1989) found during the identification of X-ray cluster sources fromthe ROSAT All-Sky Survey in the frame of the ESO key programme.The clusters range from poor nearby objects with z = 0.053 (left), a clus-ter with z = 0.188 (bottom left), to rich distant clusters with z = 0.243(bottom right). The figures have been produced from the STScI digitalscans of the UK-Schmidt IIIaJ plates with X-ray contours from the ROSATAll-Sky Survey images superposed. The contours are 1.5, 2, 3, 4, ...sigma significance contours of the X-ray signal-to-noise within aGaussian filter with a radius of 1 arcmin.

an assessment ofthe matter distribu-tion in the presentUniverse on verylarge scales ex-tending over sev-eral hundred Mpc(for a Hubble con-stant of 50 km s–1

Mpc–1). Such largescales are interest-ing for two major

22

larger scales (see also Tadros et al.,1998 for the APM cluster survey inthe optical).

X-ray astronomy has offered a uniquetool to efficiently detect and characterisegalaxy clusters out to large distances.Originating in the hot intracluster plasmathat fills the gravitational potential well ofthe clusters, the X-ray emission is anequally robust parameter for a first esti-mate of the size and mass of clusters asthe velocity dispersion measured from thegalaxy redshifts. But, while the X-ray lu-minosity can be readily detected formany clusters in an X-ray sky survey, thecollection of velocity dispersions requiresvery large and time-consuming redshiftsurveys.

In the project described here we haveembarked on a redshift survey of galaxyclusters detected in the ROSAT All-SkySurvey (Trümper 1993, Voges et al.1996), the first complete sky survey con-ducted with an imaging X-ray telescope.In the frame of an ESO key programme(Böhringer 1994, Guzzo et al. 1995) weare seeking a definite identification of allpossible cluster candidates found in theROSAT Survey in the southern celestialhemisphere above an X-ray flux limit cho-sen such as to provide a reasonably ho-mogeneous sensitivity coverage of thesky area. Within this survey programme,which we call the ROSAT ESO FluxLimited X-ray (REFLEX) Cluster Survey,we are investigating about 800 galaxyclusters and have obtained more than 300new cluster redshifts.

The present article describes methodsof the cluster detection and identificationas well as first preliminary results on thestatistics of the cluster population and aview on the large-scale distribution of theclusters. In a forthcoming issue of TheMessenger we describe in more detail thestrategy of the optical observations andthe spectroscopic results.

2. Cluster Detection and Follow-up Observations

The X-ray sky atlas constructed from theROSAT All-Sky Survey Mission with itsmore than 100,000 X-ray sources (Vogeset al. 1996) contains thousands of most-ly unidentified galaxy clusters. Since themajority of these sources are charac-terised by few photons, only the bright-est, well-extended X-ray cluster sourcesare readily identified, while the main partof the identifications has to be based onfurther optical information. For the RE-FLEX Survey we made mainly use of theCOSMOS optical object catalogue (e.g.Heydon-Dumbleton et al. 1989) pro-duced from the digital scans of the pho-tographic UK Schmidt Survey (providingstar/galaxy separation for the sky objects

with high completeness down to bj = 20.5mag.). The cluster candidates are foundas overdensities in the galaxy density dis-tribution at the position of the X-raysources (see Böhringer et al. in prepa-ration). Not all the sources flagged by thistechnique are true galaxy clusters, how-ever. The price paid for aiming at a highcompleteness in the final catalogue is alow detection threshold in the galaxydensity leading to a contamination of thecandidate list by more than 30% non-cluster sources. This large contaminationcan be reduced to about one-third by adirect inspection of the photographicplates, the detailed X-ray properties, andusing all the available literature informa-tion.

The subsequent work is the observa-tional task of the ESO key programmecomprising a total of 90 observing nightsdistributed evenly among the 1.5-m, 2.2-m, and 3.6-m telescopes at La Silla.These follow-up observations have atwo-fold purpose: we search for a definiteidentification for the unknown objectsand collect redshifts for all clusters forwhich this parameter is not known. Tomake optimal use of the ESO telescopes,we observe nearby, poor clusters andgroups in single-slit mode with the small-er telescopes and use the 3.6-m tele-scope with EFOSC in multi-slit operationto get multiple galaxy spectra of dense,rich clusters. Several coincident redshiftsstrongly support the cluster identificationand the spectroscopy of central dominantearly-type galaxies at the X-ray maximumplays a particularly important role in thisstudy.

The observing programme will becompleted end of 1998. Having observedthe clusters with higher X-ray flux withhighest priority, we can construct a firstcatalogue of clusters for the brighter partof the sample down to an X-ray fluxlimit of 3 · 10–12 erg s–1 cm–2 (in theROSAT band 0.1–2.4 keV) comprising

Figure 2: Cumulative number counts of X-ray clusters as a function of the limiting flux for theREFLEX sample.

Figure 3: Distribution of the X-ray luminosities as a function of redshift for the galaxy clustersof the REFLEX sample.

z

23

475 objects. 53% of these clusters canbe found in the main catalogues of Abell(1958) and Abell, Corwin, and Olowin(1989) and further 10% in the supple-mentary list. Most of the remaining clus-ters are previously unknown objects.This result highlights the importance of theselection process in the construction ofthe sample: a significant fraction of the X-ray bright and massive clusters wouldhave been missed had the X-ray clus-ters only been identified on the basisof existing catalogues. The fraction ofknown Abell clusters decreases further ifone goes to lower X-ray flux limits e.g. inthe extended REFLEX sample. Weshould also note here, however, thatone part of the objects missing in Abell’scompilation are nearby, X-ray brightgroups, which are not rich enough in theirgalaxy content to fulfil Abell’s criteria.

The high completeness of this sampleis demonstrated by a counter-test, inwhich we searched for X-ray emission atthe position of all Abell and ACO clustersin the ROSAT Survey independent of aprevious detection of these sources by thesurvey source identification algorithm.Only 5 clusters (~1%) were found to havebeen missed by the cluster search basedon the COSMOS data. Figure 1 gives ex-amples of three non-Abell clusters foundat different redshifts (z = 0.053, 0.188,0.243). A fraction of the X-ray sources inthe cluster candidate list are found to havenon-cluster counterparts in the follow-upobservations including AGN in clusterswhere the AGN provide the main sourceof the X-ray emission. These sourcesare removed from the sample. We alsofind a small number of cluster X-raysources in which AGN contribute to theoverall X-ray luminosity. As far as the dif-ferent sources of X-ray emission can bedistinguished by their extended and point-like nature, we have disentangled these

contributions. For an estimated verysmall fraction of sources of a few percentthis contamination may remain unde-tected.

A few spectacular discoveries weremade in the course of this programme

including the most X-ray luminous clus-ter discovered so far, RXJ1347-1145(Schindler et al. 1995, 1996) and clustersfeaturing bright gravitational arcs, e.g.S295 (Edge et al. 1994).

3. Properties of the X-rayClusters of Galaxies

In the following we are reporting resultsbased on the X-ray bright sample of 475galaxy clusters for which 413 clusterredshifts have been obtained and reducedso far or were collected from the litera-ture. A plot of the number counts of thecluster population as a function of the lim-iting X-ray flux is shown in Figure 2. Thelogarithmic graph has a slope that is veryclose to an Euclidian slope of –3/2. Thisslope is easily explained by the fact thatthe majority of the clusters are not verydistant – the peak in the redshift histogramis at z ~ 0.06 – and thus band-correctionsand evolutionary corrections are not im-portant. Figure 3 shows the distributionin redshift and X-ray luminosity. Whilemost of the clusters detected are not verydistant (median redshift is z = 0.085), afew very luminous clusters are found outto redshifts of z = 0.3 with one outstand-ing object at a redshift of z = 0.45, themost luminous cluster mentioned above.One also notes clearly the break at a red-shift of z = 0.3, which is caused by the lim-ited depth of the optical plate material

Figure 4: Galaxy group with an exceptionally dominant central cD galaxy at a redshift of z =0.0313 in the REFLEX sample. The CCD image is a short routine exposure (without filter) todefine the MOS slit mask taken with EFOSC1 at the 3.6-m telescope. The scale of the imageis about 3.5 by 3.5 arcmin.

Figure 5: Differential X-ray luminosity function for the REFLEX sample (points with error bars;for details see Böhringer et al. in preparation) compared to the previous results by DeGrandi(1996) and Ebeling et al. (1997). A value of H0 = 50 km s–1 Mpc–1 is assumed for the scaling.

24

used for the cluster pre-identification.Thus we can expect that there are more,in fact very interesting luminous clustersin the X-ray source list for this flux limitwith redshifts between z = 0.3 to 0.5. Buta more extensive imaging search pro-gramme would be necessary to findthese objects.

Another very interesting part of this X-ray cluster population is found amongthe nearby, low X-ray luminosity objects.These are groups of galaxies dominatedby giant cD galaxies that are often foundto be two or three magnitudes brighterthan the next brightest galaxy in the sys-tem. A CCD image of such a group isshown in Figure 4.

The most straightforwardly obtainedand important distribution function of thecluster sample is the X-ray luminosityfunction. This function is most closely re-lated to the mass function of the clusterswhich is used as an important calibrator

of the amplitude of the density fluctuationpower spectrum of the Universe (e.g.White et al. 1993). A preliminary versionof the X-ray luminosity function of thesample is shown in Figure 5. Note that thisfunction was derived for the sample when~ 80% of the redshifts had been obtained.But in spite of this incompleteness, theluminosity function already recovers thedensities reached in previous surveys(e.g. Ebeling et al. 1997, DeGrandi 1996)to which the present result is comparedin the figure. The high quality of the pre-sent sample is also demonstrated by thefact that the slope of the number countfunction of Figure 2 is close to the ex-pected Euclidian slope.

The most exciting aim of this pro-gramme is to assess the large-scalestructure of the galaxy cluster and mat-ter distribution in the Universe. The cur-rently most popular and important statis-tical function for the characterisation of the

large-scale structure is the density fluc-tuation power spectrum. The form of thepower spectrum depends on such basicfeatures of our Universe as the mean den-sity, the nature of the dark matter, and thephysics of the inflationary process if suchan event happened in the early Universe.Its knowledge provides therefore very im-portant constraints on the possible cos-mological models. A first preliminary re-sult for this function obtained from 188clusters of our sample in a box of 400 h–1

100Mpc is shown in Figure 6 (details inSchuecker et al., in preparation). Thefunction is featuring an interesting max-imum at a scale of about 100–150 h–1

100Mpc. The location of this maximum is re-lated to the size of the horizon of theUniverse at an epoch when the energydensity of matter and radiation was equaland it is therefore a very important cali-bration point for the theoretical modellingof the evolution of our Universe. Note thatthe construction of the power spectrumshown in Figure 6 is based on a luminosity(mass) selection of the clusters whichvaries with redshift. Therefore, the quan-titative inter-pretation of this result is notstraightforward and more work is need-ed to relate this function to the powerspectrum of the matter density fluctuationin the Universe.From the power spectrum shown wecan also derive another statistical func-tion which is more illustrative to non-ex-perts, the rms-fluctuation level on differ-ent scales as provided by a filtering of thecluster density fluctuation field by a sim-ple Gaussian smoothing filter. This func-tion is shown in Figure 7 where we notethat at the scale of the maximum of thepower spectrum (a scale of ~ 100 Mpc)the cluster density varies typically byabout 10%, while at the largest scales(400 Mpc) a fluctuation level of the orderof 1% is indicated. This sets a high stan-dard of requirements on the cluster sur-vey, and while we are confident that thepresent sample contains no systematic bi-ases on the 10% level, further tests andsimulations are needed to explore the re-liability of the results on the largestscales.

4. Conclusions

It is obvious from the above argumentsthat studies of the large-scale structureusing clusters as probes require greatcare in the preparation of the sample oftest objects. The first results of the RE-FLEX Survey presented here and somefurther tests that we have conducted al-ready demonstrate the high quality of thepresent sample. This is the result of a veryhomogeneous and highly controlled se-lection process used to construct the sam-ple. The challenge of the selection workin this survey is to combine the comple-mentary information from X-ray and op-tical wavelength in the most homoge-neous way. Contrary to several earlierstudies that we have conducted (Pierreet al. 1994, Romer et al. 1994, Ebelinget al. 1997, DeGrandi 1996) in which the

Figure 6: Power spec-trum of the density dis-tribution of the galaxyclusters in the REFLEXsample. For the analy-sis, only 188 clustersin a box of 400 Mpcside length is used(Schuecker et al. inpreparation). The re-sults are scaled to H0 =100 km s–1 Mpc–1. Thehorizontal axis givesthe wave vector ofthe Fourier compo-nents such that thecorresponding physi-cal scale is 2π/k. Thesolid line is a paramet-ric fit to the data not re-ferring to a particularcosmological model.

Figure 7: Square root of the variance of the fluctuations in the cluster density distribution as afunction of scale (obtained with Gaussian filtering). A value of H0 = 100 km s–1 Mpc–1 is assumedfor the scaling.

25

identification process was optimised tofind as many clusters as possible by us-ing all available sources of information,we have now achieved a highly completecluster selection by just combining theROSAT All-Sky Survey data and theCOSMOS optical data base in a ho-mogeneous way, completely controlled byautomated algorithms. Additional infor-mation is only used in the final identifi-cation but does not influence the selec-tion. This is a very important achievementin this survey work.

The data presented are still not fullycomplete in redshifts. But with data al-ready obtained in January and September1998 we can practically complete this dataset (to 96%). An extended sample ofREFLEX clusters down to a flux limit of2 · 10–12 erg s–1 cm–2 is already preparedand redshifts are available for more than70% of the objects. This extended set ofabout 750 galaxy clusters will help verymuch to tighten the constraints for thepower spectrum and extend it to largerscales. It will further enable us to inves-tigate the cluster correlation function – inparticular the X-ray luminosity depen-dence of the clustering amplitude, whichis an issue not yet resolved. Finally, a

complementary ROSAT Survey clusteridentification programme is being con-ducted in the Northern Sky in a collabo-ration of the Max-Planck-Institut fürExtraterrestrische Physik and J. Huchra,R. Giacconi, P. Rosati and B. McLeanwhich will soon reach a similar depth andprovide an all-sky view on the X-raycluster distribution.

References

Abell, G.O., 1958, ApJS, 3, 211.Abell, G.O., Corwin, H.G. and Olowin, R.P.,

1989, ApJS, 70, 1.Böhringer, H., 1994, in Studying the Universe

with Clusters of Galaxies, H. Böhringer andS.C. Schindler (eds.), Proceedings of an as-trophysical workshop at Schloß Ringberg,Oct. 10–15, 1993, MPE Report No. 256, p.93.

DeGrandi, S., 1996, in Röntgenstrahlung fromthe Universe, H.U. Zimmermann, J.E.Trümper, H. Yorke (eds.), MPE Report263, p. 577.

Edge, A.C., Böhringer, H., Guzzo, L., et al.,1994, A&A, 289, L34.

Ebeling, H., Edge, A.C., Fabian, A.C., Allen,S.W., Crawford, C.S., and Böhringer, H.,1997, ApJ, 479, L101.

Guzzo, L., Böhringer, H., Briel, U., et al., 1995,in Wide-Field Spectroscopy and the DistantUniverse, S.J. Maddox and A. Aragón-Salamanca (eds.), World Scientific, Singa-pore, p. 205.

Heydon-Dumbleton, N.H., Collins, C.A., andMacGillivray, H.T., 1989, MNRAS, 238,379.

Pierre, M., Böhringer, H., Ebeling, H., Voges,W., Schuecker, P., Cruddace, R.G., andMacGillivray, H.T., 1994, A&A, 290, 725.

Romer, A.K., Collins, C.A., Böhringer, H.,Cruddace, R.G., Ebeling, H., MacGillivray,H.T., and Voges, W., 1994, Nature, 372, 75.

Schindler, S.C., Guzzo, L., Ebeling, H., et al.,1995, A&A, 299, L9.

Schindler, S., Hattori, M., Neumann, D.M., andBöhringer, H., 1996, A&A, 317, 646.

Tadros, H., Efstathiou, G., and Dalton, G.,1998, MNRAS, 296, 995.

Trümper, J., 1993, Science, 260, 1769.Voges, W., Boller, T., Dennerl, K., et al.,

1996, in Röntgenstrahlung from theUniverse, H.U. Zimmermann, J.E. Trümper,H. Yorke (eds.), MPE Report No. 263, p.637.

White, S.D.M., Efstathiou, G., and Frenk,C.S., MNRAS, 1993, 262, 1023.

hxb@mpe.mpg.de

Timing, Spectroscopy and Multicolour Imaging ofthe Candidate Optical Counterpart of PSR B1509–58R.P. MIGNANI 1, S. MEREGHETTI 2, C. GOUIFFES 3 and P. A. CARAVEO 2, 4

1ST-ECF, Garching, rmignani@eso.org 2Istituto di Fisica Cosmica “G.Occhialini”, Milan 3Service d’Astrophysique CEA/DSM/DAPNIA C.E Saclay 2,4Istituto Astronomico, Università La Sapienza, Rome, Italy

1. Introduction

Optical counterparts have now beenproposed for nine Isolated Neutron Stars(INSs). For some of them (the Crab andVela pulsars, PSR B0540–69) the iden-tifications have been confirmed throughthe detection of optical pulsations (with atentative detection existing also for PSRB0656+14) or, in the case of Geminga,initially from the proper motion of the pro-posed counterpart. For the rest of thesample (PSR B1509–58, PSR B1055–52,PSR B1929+10 and PSR B0950+08), theoptical identification still relies on the po-sitional coincidence with a field object (seee.g. Caraveo 1998 and Mignani 1998 fora summary). Unfortunately, in most cas-es the intrinsic faintness of such objectshampers the timing of their optical emis-sion, thus making necessary fast-pho-tometry facilities attached to 4-m-classtelescopes.

Among the uncertain cases, the beststudied is certainly PSR B1509–58. Witha dynamical age close to 1500 yrs, PSRB1509–58 is the youngest INS after the

Crab. While its period (P=150 ms) is longcompared with that of the similarly oldCrab pulsar and PSR B0540–69, its spindown rate P

.(p 1.5 1012s s–1) is the high-

est in the pulsar family. This made it pos-sible to obtain an accurate measurementof the P

..and thus of the pulsar braking in-

dex (Kaspi et al. 1994). Since PSRB1509–58 lies close to the geometricalcentre of the plerionic supernova remnantMSH15–52 (Strom 1994), it may be oneof the very few cases of a pulsar/plerionassociation. However, the ages of the pul-sar and of the remnant are significantlydifferent (Gaensler et al. 1998), castingdoubts on the association.

PSR B1509-58 was first detected in X-rays by the Einstein Observatory (Seward& Hardnen 1982) and soon after in radio(Manchester, Tuhoy and D’Amico 1982)with a single pulse profile preceding inphase the broad, asymmetric, X-raypeak. Pulsations in the 90–600 keVrange have been detected by BATSE(Matz et al. 1994) and OSSE (Ulmer etal. 1993) on board GRO while only an up-per limit on the source flux at E M 100

MeV was obtained with EGRET (Brazieret al. 1994).

In the optical, a candidate counterpart(V p 22), coincident with the pulsar co-ordinates reported by Taylor, Manchesterand Lyne (1993) – TML93 – was pro-posed by Caraveo et al. (1994a). Werethis object indeed the pulsar, at a distanceof 4.4 kpc (TML93), its optical emissionwould certainly be magnetospheric, whichis also expected from its young age.However, the corresponding luminosityexceeds by a few orders of magnitude thevalue expected on the basis of the PaciniLaw (Pacini 1971) i.e. Lopt ∝ B4P–10

(where B is the pulsar magnetic field)which works for the other young pulsars(Pacini & Salvati 1987). A firm confirma-tion of the optical identification is thus oforder. Of course, searching for optical pul-sations at the radio period is the defaultway.

The results of timing of the candidatecounterpart were first reported byCaraveo (1998) and soon after con-firmed by the independent works ofChakrabarty & Kaspi (1998) and of

•

26

Shearer et al (1998). In all cases, no op-tical pulsations at the radio period wereobserved, thus leaving the problem of theidentification open.

In the next sections we will discuss theresults of detailed optical investigationsof the PSR B1509–58 candidate coun-terpart performed by our group with theESO telescopes. Apart from the fastphotometry observations of Caraveo(1998), presented here in detail, the dataset consists of both multicolour imagingand spectroscopy.

2. Data Overview

2.1 Timing

Optical timing of the Caraveo et al.counterpart was performed in May 1994from the ESO 3.6-m telescope (Cara-veo et al. 1994b). The telescope wasequipped with the ESO fast (0.1 ms timeresolution) photometer, a single-channeldevice with a GaAs photocatode sensi-tive in the range 3500–9000 Å, mainlyused to search for an optical pulsar inSN1987A but also for the timing of PSRB0540–69 (Gouiffes, Finley and Ögelman1992) and of the Vela Pulsar (Gouiffes1998). A total of 9 observations, split intwo adjacent nights, were performed(see Table 1) with two different filters i.e.a standard R and the og590 which cutsat wavelengths shorter than 5800 Å.According to the seeing conditions, twodifferent apertures were used with diam-eters of 7 and 4 arcsec, respectively.

Photon arrival times have been cor-rected for the Earth rotation and revolu-tion (barycentric correction) using theJPL/DE200 FORTRAN code. The peri-odicity search was performed applying thestandard folding technique. First, thewhole procedure was tested using as areference the p 5000-s observation of theslightly fainter (V p 22.5) PSR B0540–69(see Fig.1). The procedure was then ap-plied to the data of PSR B1509–58. Dueto the better seeing and atmospheric con-ditions, we decided to concentrate ouranalysis on data taken the second night.For each observation, photon countshave been resampled in bins of 4 and 8

ms. The time series have been folded atdifferent trial periods stepping ± 10–8 sfrom the expected period, computed us-ing the ephemeris of Kaspi et al. (1994).For each trial period the corresponding χ2

has thus been computed but no signifi-cant maximum was observed in the χ2 dis-tribution. The periodicity search was re-peated using the whole data set of thesecond night (Q 20,000 s) but no signif-icant improvement was achieved. Usingas a zero point the observation of a stan-dard star (V = 16.125) from an E5 regionof a Graham field, we could derive an up-

per limit R p 23–23.5 on the pulsed mag-nitude, which corresponds to a pulsedfraction M 15%, similar to the value giv-en by Chakrabarty & Kaspi (1998).

This means that if the optical candidateof Caraveo et al. is indeed the counter-part of PSR B1509–58, its pulsed fractionmust be much smaller than for the Craband Vela pulsar, which have pulsed frac-tions p 100% and M 50%, respectively.

2.2 Spectroscopy and Imaging

To investigate the possibility of a wrongassociation with a fore/background object,in June 1995 we performed spectroscopyof the object from the NTT. Three 90-minute, medium-resolution (2.3 Å/pixel)spectra were obtained with EMMI-Red(ESO Multi Mode Instrument) in thewavelength range 3800–8400 Å. Unfor-tunately, the very low S/N made it im-possible to perform any spectral analysisof the source, apart from excluding thepresence of strong emission/absorptionlines.

Multicolour images were also collect-ed from the NTT during different observ-ing runs (see Table 2 for a summary). Tocomplement and improve the first mea-sures in V and R of Caraveo et al.(1994a) we collected longer exposureswith SUSI in B, V and the first exposurein I in 1995. A second deep observation

Obs.# MJD T (s) r (arcsec) CR (c/s) Filt Obj

1 49478 1468 4 756 og590 PSR B1509-582 2821 4 825 og590 PSR B1509-583 462 7 2944 og590 PSR B1509-584 6305 7 3269 og590 PSR B1509-585 6097 7 1267 R PSR B1509-58

6 49479 5031 4 1687 free PSR B0540-697 3616 4 305 R PSR B1509-588 2745 4 312 R PSR B1509-589 6809 4 294 R PSR B1509-5810 6813 4 316 R PSR B1509-5811 4094 4 640 R Graham std.

Table 1: Journal of fast photometry observations. Columns list the observing sequence, the ob-serving epochs (MJDs), the duration of the single observations (s), the aperture diameter (arc-sec), the count rate in the aperture (cts/s), the filter name and the target identifier.

Figure 1: Light curve (Q 5 ms/bin) of PSR B0540–69 folded over the expected period (Pexp) com-puted by time-propagating at the observing epoch the pulsar’s P

.– P..

according to the ephemerisof Gouiffes, Finley and Ögelman (1992). Two cycles are shown for clarity. A single broad peakis clearly visible in the light curve χ 2 Q 50), exactly as expected from previous fast photometryobservations. The clear detection of the pulsar shows both the sensibility of the instrument andthe correctness of our procedure.

27

in R was obtained in 1997. After cosmic-ray rejection, bias subtraction and flatfielding, a zero point was computed foreach filter using photometric standards inthe Stobie and Landolt fields. These newobservations yielded more accurate V(22.1 ± 0.1) and R (20.8 ± 0.1) magnitudestogether with the first measurements inB (23.8 ± 0.3) and I (19.8 ± 0.1). In orderto assess the colours of the object, an un-known but potentially important amountof interstellar absorption should be con-sidered. X-ray spectral fittings give an NHQ 8 × 1021 cm–2 (Becker & Trümper 1997)with an, at least, 50% uncertainty whichimplies a difference of about one magni-tude on the colour indexes. Assuming thatthe object be at the pulsar distance of 4.4kpc, it could be thus anything from a Gstar (for the lowest value of the NH ) to amuch highly absorbed O or B star. Tosearch for other possible candidates wehave used our deepest exposure of thefield (the 1997 one) taken with SUSI un-der better seeing conditions ( p 0.7 arc-sec). The revised radio co-ordinates ofPSR B1509–58, reported in Taylor et al.(1995) have been registered on the im-age using the UK STARLINK softwareASTROM (Wallace 1990) and a referenceframe given by several USNO stars iden-tified in the SUSI field (2 × 2 arcmin).Adding in quadrature to the pulsar posi-tional error ( 1.3 arcsec) an average un-certainty of 0.25 arcsec attached to theabsolute position of the USNO stars, andthe r.m.s. of the plate solution fit (0.42 arc-sec) we can estimate a conservative er-ror of 1.4 arcsec on our astrometry. Theradio position of the pulsar is thus shownin Figure 2, marked by a yellow cross.As already noted by Caraveo (1998),the revised radio co-ordinates of PSRB1509–58 are slightly different from theold ones reported in TML93, also markedin Figure 2 (in blue). However, the asso-ciated errors (w 1 arcsec in both dimen-sions) leave room for a substantial agree-ment. The object inside the overlappingerror circles is the candidate counterpartof Caraveo et al. (1994a).

As evident from Fig.2, no other sourceis visible inside the uncertainty regions,down to a lower limit of R p 25.

3. Conclusions

No pulsations were detected from theproposed optical counterpart of PSRB1509-58, thus adding more weight to theresults obtained by Chakrabarty & Kaspi(1998) and Shearer et al. (1998). Unless

invoking a peculiar-emission geometry,the lack of optical pulsations strongly ar-gues against the identification of Caraveoet al. (1994a). Given the high interstellarabsorption towards the pulsar, multi-colour imaging does not help to constrainthe real nature of the object. Spectros-copy, to be performed with the UT1 of theVLT, is thus needed. Were the opticalidentification of Caraveo et al. be dis-proved, the real counterpart of PSRB1509-58 is to be searched somewherein the close surroundings. In this case, itmust be fainter than R p 25. This is inagreement with the Pacini law (1971)which would predict for PSR B1509-58 amagnitude p 27.

Although this values is certainly with-in reach of the VLT (Mignani et al. 1999),if located a fraction of arcsec from thepresent candidate, such a faint objectwould be hardly visible from the groundeven under exceptional seeing condi-tions. The situation would thus be simi-lar to the case of PSR B1055-52, close(m 4 arcsec) to a p 14.6 magnitudes

Figure. 2: 16 ×16 arcsec2 R-band image of the field of PSR B1509–58, taken in June 1997 withthe ESO/NTT. The image has been obtained with the SUperb Seeing Imager (SUSI) under sub-arcsec seeing conditions (Q 0.7 arcsec). The exposure time was 45 min. The pixel size of theCCD is 0.13 arcsec. The yellow cross marks the position of PSR B1509–58 computed accord-ing to the revised radio coordinates of Taylor et al (1995). The circle (r = 1.4 arcsec) correspondsto the uncertainty region associated to the pulsar position, resulting from the combination of theerror on the radio coordinates (r p 1.3 arcsec), the average uncertainty on the absolute coor-dinates of the USNO stars (p 0.25 arcsec) and the σ of the astrometric fit (p 0.42 arcsec). Thecandidate optical counterpart of the pulsar proposed by Caraveo et al. (1994a) is visible insidethe error circle. For completeness, the old pulsar position from TML93 (r Q 0.9 arcsec) is alsoshown in blue. The faint object (R Q 24) seen about 2 arcsec north of the pulsar position andmarked by the arrow is too far to claim any association.

Date Det Filt. T mag

July 93 EMMI-Red V 5m 22.0 w 0.2EMMI-Red R 10m 20.8 w 0.3EMMI-Red B 5m O 23

Jan 95 SUSI B 20m 23.8 w 0.3SUSI V 25m 22.1 w 0.1SUSI I 5m 19.8 w 0.1

Jul 97 SUSI R 45m 20.8 w 0.1

Table 2: Summary of the available photometry of the PSR B1509–58 candidate counterpart per-formed with the NTT from 1993 to 1997. Columns list the observing epochs, the detector used,the filters, the exposure times and the observed (not-dereddened) magnitudes. The July 1993observations are the ones also reported in Caraveo et al. (1994a).

28

brighter field star and unresolved by theNTT/SUSI (Mignani et al. 1997). As in thatcase, high-resolution imaging of the re-gion with the HST would be the only wayto pinpoint the pulsar.

Acknowledgements

We acknowledge the support soft-ware provided by the Starlink Project,which is funded by the UK SERC. C.Gouiffes greatly acknowledges the sup-port of the technical staff at La Silla forhis invaluable and constant help in the useof the fast photometer.

References

Becker, W. & Trümper, 1997, A&A 326, 682.Brazier, K.T.S. et al, 1994, MNRAS, 268, 517.

Caraveo, P.A., Mereghetti, S. and Bignami,G.F., 1994a, ApJ 423, L125.

Caraveo, P.A., Bignami, G.F., Mereghetti, S.,Mignani, R. and Gouiffes, C., 1994b, BAAS184, 2104C.

Caraveo, P.A., 1998, Adv. in Sp. Res. Vol. 21,187.

Chakrabarty, D. & Kaspi, V., 1998, ApJ 498,L37.

Gaensler, B.M., Brazier, K.T.S., Manchester,R.N., Johnston, S. and Green, A.J., 1998,Memorie della Società Astronomica Italiana– in press.

Gouiffes, C., Finley, J.P. and Ögelman, H.,1992, ApJ 394, 581.

Gouiffes, C., 1998, Neutron Stars and Pul-sars: Thirty years after the discovery.ed. N. Shibazaki, N. Kawai, S. Shibata andT. Kifune, p. 363, Universal AcademicPress.

Kaspi, V.M. et al., 1994, ApJ 422, L83.Manchester, R.N., Tuhoy, I.R. and D’Amico,

N., 1982, ApJ 262, L31.

Matz, S.M. et al, 1994, ApJ 434, 288.Mignani, R., Caraveo, P.A. and Bignami, G.F.,

1997, ApJ 474, L51.Mignani, R., 1998, Neutron Stars and Pulsars:

Thirty years after the discovery. ed. N.Shibazaki, N. Kawai, S. Shibata & T. Kifune,p335, Universal Academic Press.

Mignani, R., Caraveo, P.A. and Bignami, G.F.,1999, A&A – in press.

Pacini, F., 1971, ApJ 163, L17.Pacini, F. & Salvati, M., 1987, ApJ 321, 447.Seward, F.D. &Hardnen, F.R., 1982, ApJ 256,

L45.Shearer, A. et al, 1998, A&A 333, L16.Strom, R.G., 1994, MNRAS 268, L5.Taylor, J. H., Manchester, R. N. and Lyne, A.

G., 1993, ApJS 88, 529.Taylor, J.H., Manchester, R.N., Lyne, A.G. and

Camilo, F.,1995 (http://pulsar.princeton.edu/pulsar/catalog.shtml)

Ulmer, M.P. et al, 1993, ApJ 417, 738.Wallace, P.T., 1990, Starlink User Note 5.11,

Rutherford Appleton Laboratory.

The increasing number of browndwarfs discovered in the last few yearsis rapidly opening the possibilities ofstudying a wide range of their propertiesand the ways in which these depend onessential parameters, such as the mass,the age, the rotation, or the environment.One of these properties is the magneticfield, which in principle should be ex-pected to be important in fully convectiveobjects such as brown dwarfs. The chro-mospheric X-ray emission, widely ob-served in M-type dwarfs (Neuhäuser1997), has its origin in this magnetic ac-tivity. As such, it offers an observationaltool to probe the interior of these objects,the mechanisms for the generation andmaintenance of their magnetic fields,and the way in which the magnetic activityis affected by the basic parameters of theobject. The detection of X-ray emissionfrom brown dwarfs is thus of great im-portance to extend our understanding ofthe properties of stellar magnetic fields tothe substellar domain, as well as to as-certain to what extent a small, substellar

mass, and the consequent lack of a per-manent nuclear energy source, can havean impact in the production and the evo-lution of a magnetic field.

Until recently, no conclusive evidencefor X-ray emission from brown dwarfs had

been found, as shown by an extensivesearch in ROSAT archive observationsnear the position of known bona-fideand candidate brown dwarfs (Neuhäuseret al. 1999). However, a newly identifiedmember of the Chamaeleon I star form-

The First X-ray Emitting Brown DwarfF. COMERÓN, ESO, Garching, GermanyR. NEUHÄUSER, MPE, Garching, GermanyA.A. KAAS, Stockholm Observatory, Stockholm, Sweden

Figure 1: A 13′ ×13′ image of the centre of theChamaeleon I aggregate in the I band. Labelsidentify previously known members of the ag-gregate, and numbers 1 to 6 denote the low-mass members newly found in the Hα surveyby Comerón et al. 1999, i.e. Cha Hα 1 to 6.This image was obtained using DFOSC at the1.5-m Danish telescope on La Silla, and in-cludes the area surveyed in the infrared us-ing IRAC2b at the ESO-MPI 2.2-m telescope.The area covered by the X-ray and ISOCAMobservations is considerably larger.

ing cloud, whose temperature and lumi-nosity clearly identify it as a brown dwarfyounger than one million years, is clear-ly detected by a deep ROSAT observa-tion with a 9σ detection significance(Neuhäuser & Comerón 1998). This ob-ject is only the second bona-fide browndwarf to be detected in a star-forming re-gion (Luhman et al. 1997), and the firstone known to emit in X-rays.

The identification of the X-ray emittingbrown dwarf in Chamaeleon I, hereafterreferred to as Cha Hα 1, is the result ofthree parallel efforts aimed at findingbrown dwarfs in this star-forming regionby means of different characteristics typ-ical of low-mass young stellar objects. Thefirst of them was a research on deep ob-servations in the ROSAT archive,amounting to 37.8 ksec of effective ob-serving time at the position of Cha Hα 1,looking for faint, previously unidentifiedChamaeleon I members revealed bytheir X-ray emission. The second one con-sisted of the combination of a JHK sur-vey of 100 arcmin2 of the centre of the ag-gregate, carried out with IRAC-2b at theESO-MPI 2.2-m telescope on La Silla inMay 1997, and an objective-prism surveyof approximately the same field usingDFOSC at the 1.5-m Danish telescope,also on La Silla, in March 1998. The aimin this case was to identify new faint mem-bers by means of their possible near-in-frared excess or by their emission in Hα.A third set of observations came from alarge-area star-formation survey carriedout with ISOCAM at 6.7 and 14.3 µm(Nordh et al. 1996). The aim of this sur-vey was to identify young stellar objectsby means of their mid-IR excess in thecolour index [14.3/6.7] and obtain a morecomplete census of the young stellar pop-ulation for IMF studies. From an empiri-cal correlation between the flux at 6.7 µmof the excess sources and bolometric lu-

minosities (Prusti et al. 1992) extrapolatedto faint sources, as well as adopting themedian age of T Tauri stars of 3 106 yr(Lawson et al. 1996) as a common agefor these, the theoretical models ofD’Antona & Mazzitelli (1997) predict thatabout 25% of the ISOCAM selected

young stellar objects in Cha I are sub-stellar (Olofsson et al. 1998), a result thatmust be confirmed spectroscopically.

Cha Hα 1 is identified in all these sur-veys. Figure 1 shows it in an I-band im-age of the field surveyed in Hα, obtainedalso with the 1.5-m Danish telescope.Also indicated are the previously knownmembers of the aggregate (e.g. Lawsonet al. 1996), as well as five other newmembers identified by their Hα emissionwhose masses are near, and possibly be-low, the limit separating stars from browndwarfs (Comerón et al. 1999). The X-rayimage of the same field is shown inFigure 2. The initial spectrum of Cha Hα1 from the DFOSC survey revealed aprominent emission at Hα and a late Mspectral type, but given the faintness ofthe object in the Hα region, it was not pos-sible to precisely determine the spectralsubtype, an essential element to ascer-tain its temperature and therefore itsmass. For this reason, additional spec-troscopic observations at visible andnear-infrared wavelengths were carriedout in May 1998 using EMMI and SOFIat the NTT.

The visible spectrum of Cha Hα 1 ob-tained with the NTT is shown in Figure 3,where it is compared to late M spectralstandard stars. The overall appearanceof the spectrum clearly indicates a late Mtype, as confirmed as well by several in-dicators which measure the strengths oftemperature-sensitive spectral features.

29

Figure 2: Deep ROSAT X-ray image of the centre of the Chamaeleon I aggregate, covering anarea similar to that of Figure 1. The position of all the known and newly identified members isindicated.

Figure 3: The visible spectrum of Cha Hα 1 obtained with EMMI at the NTT. The M6V com-parison spectrum is that of Gl 406; M7V, of vB 8; M8V is the combination of LHS 2243, LHS2397 A, LP 412-31, and vB 10; and M9V, of BRI 1222–1222, LHS 2065, LHS 2924, and TVLM868–110639. The spectral classification of the standard stars is from Kirkpatrick et al. 1995.

Based on these elements, we classifyCha Hα 1 as M7.5–M8. Some differenceswith the standard field M dwarfs are ap-parent as well: the principal one is thesmaller depth of the Na I absorption fea-ture at 8190 Å, which is known to be sen-sitive to gravity. The smaller depth is infact a confirmation of the youth of Cha Hα1, as an object in the early stages of con-traction should have a surface gravity in-termediate between that of a dwarf (whereNa I is strong) and a giant (where Na I isweak).

The precise spectral classification ob-tained from the NTT visible spectrummakes it possible to derive the surfacetemperature of Cha Hα 1, using existingspectral type-temperature calibrations(Leggett et al. 1996). In this way, we ob-tain a temperature of 2600 K. On the oth-er hand, the luminosity can be determinedfrom the visible and near-infrared pho-tometry obtained with our DFOSC andIRAC2b observations respectively: weadopt for Cha Hα 1 the distance of 160pc to Chamaeleon I recently derivedfrom Hipparcos data and extrapolate toM7.5 the bolometric corrections and in-trinsic colours of Kenyon and Hartmann(1995) in order to find the reddening-cor-rected luminosity of the object. In this way,we obtain logL(L0) = –1.79. The derivedtemperature and luminosity allow the di-rect comparison with the results of pre-main-sequence evolutionary tracks, thusallowing an estimate of the age and

mass of Cha Hα 1. Using pre-main-se-quence tracks by Burrows et al. (1997),we obtain a mass of 0.03 M0 and an ageof 5 × 105 years. The position of Cha Hα1 in the H-R diagram, and the lines ofequal mass and age in that diagram pre-dicted by the isochrones, are shown inFigure 4. The results are essentially un-changed by using tracks by D’Antona &Mazzitelli (1997), whose main differenceis the allocation of a somewhat youngerage for our objects, p 3 × 105 years.Strictly, the temperature scale adopted isvalid only for field dwarfs, whose surfacegravities are higher; M giants of thesame spectral subtype tend to be hotter,what may imply that our temperature forCha Hα 1 is actually an underestimate.However, even allowing for a total un-derestimate of 200 K which would includeerrors due to surface gravity effects anduncertainties in the temperature calibra-tions through model-atmosphere fitting,Cha Hα 1 still remains safely in the sub-stellar realm, with a mass below 0.06 M0.

Cha Hα 1 displays most of the typicalsignatures of youth commonly associat-ed with low-mass young stellar objects.The X-ray luminosity derived from theROSAT observations is log LX(erg s–1) =28.3 ± 0.1, implying an X-ray to bolometricluminosity ratio of log(LX/Lbol) = –3.44, wellwithin the range occupied by low-massstars in young clusters. This result sug-gests that there is no essential differencein the X-ray emission properties of very

young stellar objects when crossing thethreshold separating stars from browndwarfs. The prominent Hα emission, withan equivalent width of 59 Å, is also typi-cal of classical T Tauri stars. On the oth-er hand, and unlike in the case of high-er-mass members of Chamaeleon I andother young aggregates, this strong Hαemission is not accompanied by a no-ticeable excess emission in the K near-infrared band; excess emission is nev-ertheless observed at 6.7 and 14.3 µm,as revealed by the ISOCAM measure-ments. The ISOCAM colour index[14.3/6.7] = 0.15 corresponds to αISO =–0.55 (α = d log(λFλ)/d log λ), which is inagreement with it being a Class II object.This colour is in rough agreement withtheoretical models of flared circumstellardisks (Kenyon & Hartmann 1995). Excessis also found in the K – m(6.7µm) colourindex. In principle, the lack of excess inthe K band is to be expected: neither vis-cous heating at the presumably low ratesof accretion on a very low mass object likeCha Hα 1, nor the irradiation by such acool central star, should produce in a cir-cumstellar disk any significant emissionin the K band, although such excess emis-sion could become important at longerwavelengths. However, we note that oth-er very low mass objects with compara-bly low temperatures do display emissionof circumstellar origin in the K-band (Co-merón et al. 1998, Wilking et al. 1998).Our discovery of X-ray emission from ChaHα 1, while revealing brown dwarfs as anew class of X-ray emitting objects,seems to suggest that the transition fromthe stellar to the substellar domain doesnot imply a dramatic change in the prop-erties of very young low-mass, fully con-vective objects. This seems to be rein-forced by the very recent identification ofX-ray emission from GY 202 (Neuhäuseret al. 1999), one of the probable browndwarfs found in the ρ Ophiuchi cluster byComerón et al. (1998 and referencestherein), whose substellar character hasreceived recent support by spectroscop-ic observations by Wilking et al. (1998).The characteristics of this object (mass,age, log(LX/Lbol)) seem to be very similarto those of Cha Hα 1. The magnetic prop-erties of massive young brown dwarfsmay therefore be simply an extension ofthose already well studied in their moremassive counterparts, M-dwarf stars.Nevertheless, a number of questions re-main open: why no X-ray emission is seenfrom other, more evolved brown dwarfs,despite the availability of observations(Neuhäuser et al. 1999) that put rather se-vere limits on the log(LX/Lbol) ratios ofsome of them? Is X-ray emission a fea-ture restricted to the very early stages ofthe evolution of brown dwarfs, perhapsdecreasing at later times as a conse-quence of the lack of a stable source ofenergy in their interiors? What is the de-pendence of X-ray emission of browndwarfs on parameters such as the mass,the age, the luminosity, the rotation, or theexistence of circumstellar material? Moredetailed studies of objects such as Cha

30

Figure 4: H-R diagram showing the position of Cha Hα 1. The lines give the position of theisochrones (solid lines) and lines of equal mass (dotted lines) for late M spectral types accord-ing to Burrows et al. 1997. The line labelled as 0.078 M0, separates brown dwarfs from lowmass stars.

Hα 1, and better constraints on the X-rayemission of other brown dwarfs, will benecessary to answer these questions.

Acknowledgements: We thank thestaff at ESO-La Silla for assistance withthe observations, and Dr. Kevin Luhmanfor the comparison spectra presented inFigure 3. This work was supported byNATO Collaborative Research GrantCRG 940671. RN wishes to thank theDeutsche Forschungs-Gemeinschaft forfinancial support in their Star-Formationprogramme. AAK thanks for the financialsupport from the Swedish NationalSpaceboard. ROSAT is supported by theMax-Planck Society and the GermanGovernment (BMBF/DLR). ISO is anESA project with instruments funded byESA Member States and with the partic-ipation of ISAS and NASA.

References

Burrows, A., Marley, M., Hubbard, W.B.,Lunine, J.I., Guillot, T., Saumon, D.,Freedman, R., Sudarsky, D., Sharp, C.,1997, ApJ, 491, 856.

Comerón, F., Rieke, G.H., Claes, P., Torra, J.,Laureijs, R.J., 1998, A&A, 335, 522.

Comerón, F., Rieke, G.H., Neuhäuser, R.,1999, A&A, in press.

D’Antona, F., Mazzitelli, I., 1997, Mem. S. A.It., 68, 807.

Kenyon, S.J., Hartmann, L., 1995, ApJS, 101,117.

Kirkpatrick, J.D., Henry, T.J., Simons, D.A.,1995, AJ, 109, 797.

Lawson, W.A., Feigelson, E.D.,Huenemoerder, D.P., 1996, MNRAS, 280,1071.

Leggett, S.K., Allard, F., Berriman, G., Dahn,C.C., Hauschildt, P.H., 1996, ApJS, 104,117.

Luhman, K.L., Liebert, J., Rieke, G.H., 1997,ApJ, 489, L165. Neuhäuser, R., 1997,Science, 276, 1363.

Neuhäuser R., Comerón F., 1998, Science,282, 83.

Neuhäuser R., Briceno, C., Comerón, F.,Hearty, T., Martín, E.L., Schmitt, J.H.M.M.,Stelzer, B., Supper, R., Voges, W.,Zinnecker, H., A&A, submitted.

Nordh, L., Olofsson, G., Abergel, A., et al.,1996, A&A, 315, L185.

Olofsson, G., Kaas, A.A., Nordh, L., et al., 1998,In: Brown Dwarfs and Extrasolar Planets,eds. Rebolo R., Martín, E.L., ZapateroOsorio, M.R., ASP Conf. Ser. 134, p. 81.

Prusti, T., Whittet, D.C.B., Wesselius, P.R.,1992, MNRAS, 254, 361.

Wilking, B.A., Greene, T.P., Meyer, M.R.,1998, AJ, in press.

31

fcomeron@eso.org

CALL FOR IDEAS FOR FUTURE PUBLIC IMAGING SURVEYS

Based on the experience gained with EIS, and to prepare for the full exploitation of future imaging facilities such as theVST, ESO is ready to co-ordinate efforts to provide the community with imaging survey data taken at the La Silla Telescopes(2.2-m and NTT).

To this end, and following the recommendation of the Working Group for Surveys (WGS), ESO is issuing the presentCALL FOR IDEAS for future public imaging surveys.

Scientists in the community are invited to fill out the form available in the ESO EIS Web page with their suggestions forfuture surveys (http://www.eso.org/science/eis/eis_ideas/wfi2.2_form.html). The compiled responses will be forwarded tothe chairman of the WGS (Joachim Krautter, J.Krautter@lsw.uni-heidelberg.de).

DEADLINE FOR SUBMISSION: FEBRUARY 15, 1999

In elaborating their suggestions interested scientists should consider the following guidelines.A Public survey should primarily be aimed at the maximisation of the scientific outcome of the VLT, and should consist

of observations not likely to be covered by “private” proposals.Public surveys should also have a broad scientific goal, provide data that could be fruitfully used by astronomers work-

ing in different astronomical areas, e.g. solar system, stellar populations, observational cosmology, etc., and should be eas-ily complemented by ‘private’ proposals aimed at more specific scientific goals. As a consequence, in general, public sur-veys may concentrate on broad-band imaging, leaving e.g. narrow-band follow up or second-epoch observations to ‘private’proposals.

The Working Group will collect the submitted ideas, and elaborate one or more detailed proposals to be submitted to theOPC, including a description of the survey products, distribution policy and schedule.

It is expected that a Survey Team, following the model of the EIS team, will be charged to execute the survey, combin-ing in it ESO scientists and Visiting scientists from the community. In the future, as standard survey tools become availablethroughout the community, one may also envisage public surveys being conducted by Teams assembled in a member stateinstitute with demonstrated experience and required infrastructure. Shortly after OPC approval of a survey, ESO will issuean announcement of opportunity for the recruitment of the Visiting Scientists who will participate in the effort.

Please read also the EIS Working Group Recommendations regarding future surveys.

EUROPEAN SOUTHERN OBSERVATORY

ATLAS OF THE NORTHERN SKYIn 1990 the Palomar Observatory (California Institute of Technology) and the European Southern Observatory (ESO) started the joint pro-

duction of the Palomar Observatory - European Southern Observatory Atlas of the Northern Sky, also called the POSS II SURVEY. The Atlas is based on photographic plates obtained with the Oschin Schmidt Telescope on Mount Palomar. It consists of 894 fields each

in Blue (B), Red (R) and near Infrared (I), covering the entire northern sky. The film copies are produced at ESO/Garching. The first copieswere made in 1990 and the Survey will be completed by the end of 2001.

We now have a few sets available for sale, either complete sets in all three colours or one-colour-sets. The amount of available sets islimited and there will be no further edition of the Atlas produced.

For more information, prices and orders please contact Christine Telander, ctelande@eso.org. Technical information can be obtained from Jean Quebatte, jquebatt@eso.org.

EUROPEAN SOUTHERN OBSERVATORYKarl-Schwarzschild-Str. 2, D-85748 Garching bei München, GermanyTel: +49-89-32006-0 – Fax: +49-89-32006-480 – http://www.eso.org

32

OTHER ASTRONOMICAL NEWS

Observations with Adaptive Optics: Getting ThereA REPORT ON THE ESO/OSA CONFERENCE, SONTHOFEN, 7-11 SEPTEMBER 1998

M.-H. ULRICH, ESO

Adaptive Optics on Large Telescopeshas opened up the new field of ground-based IR and optical astronomy on scaleof less than 0.5″. The resolution routine-ly achieved is better than 0.15″ – asif the observer had jumped more than 3times closer to the observed object – adream! Advances are occurring in allfields, from the solar system to high zgalaxies.

The first Conference centred on as-trophysical results obtained with AdaptiveOptics (AO) “Astronomy with AdaptiveOptics: Present Results and FuturePrograms” was organised jointly by ESOand the Optical Society of America andtook place 7–11 September 1998, in theBavarian resort of Sonthofen.

Here, I highlight the points that ap-peared to me as the most pressing andimportant of those discussed in and alsoout of sessions at the Conference.

These points revolve around the cen-tral question on how to use AO most ef-fectively and implement the powerfultools needed to proceed efficiently fromraw data to astrophysical papers.

To make the reading of this reporteasier, I do not give references to any ofthe papers presented at the Conferenceexcept for the figures. Full versions of thepapers can be found in the proceedings,which are now available on the Web andwill appear in print in early 1999 (seeSection 6 for Web address).

1. How to Make AO ObservationsEfficient

In astrophysics, the scientific value ofthe output of any instrument is the com-bination of the astrophysical questionsdriving the observing programmes andthe number and technical quality of thepublished papers.

Leaving aside for a moment the ques-tion of the scientific driver, one can askthe question: What is the most efficientway to obtain astrophysical results withAdaptive Optics now that this techniquehas been shown to work?

The best way is to have goals thatare in line with what the instrument candeliver – a trivial statement when ap-plied to decades-old techniques suchas direct imaging and spectrography. Itimplies, in fact, having a full understand-ing of the AO data acquisition, calibrationand analysis process at the time of theproposal preparation since this process

defines the achievable limits of perfor-mance.

The AO observations have to be un-derstood as an ensemble formed bythe simulations, the planning of thedata reduction, the scientific tactical as-pects of the optimisation at the tele-scope, the data acquisition itself, and fi-nally the data processing, analysis andpublication. The proposal preparationmust aim at an effective coupling of thisentire procedure with sound astrophysi-cal objectives.

It follows that the observer must be-come familiar with the various parts of thisensemble and have at his disposal:

– a powerful system of simulationwhere one can choose as input (i) a va-riety of real or modelled science objects,(ii) different values of the parameterswhich characterise AO performance i.e.r0, τ0, and θ0 (isoplanicity) relevant tothe Observatory, and (iii) different mag-nitudes and radial distances of the refer-ence star.

– procedures to calculate the PointSpread Function (PSF) and its noisefrom the values in (ii) and (iii) plus the re-quired exposure time needed to achievea given S/N on the science object.

– a library of deconvolution algorithmswith illustrative examples of their appli-cation.

– an observatory data bank with rawand processed data with calibrations andPSF with which the observer can acquirea first-hand knowledge of the potentialand limits of the AO instrument.

– examples of successful data acqui-sition and processing.

– catalogues of artefacts and spuriouseffects.

Moreover, at the telescope, the visit-ing observer should have some pos-sibility to modify a pre-registered ob-serving plan in order to (1) adapt tochanging seeing conditions since theperformance of AO is so dependent onthem (in some conditions, the Strehl ra-tio depends on the seeing to the sixthpower), and (2) immediately repeat orexpand observations for which an on-line quick look reveals exciting/unex-pected results – it is especially importantto have this possibility in a new field suchas high angular resolution because theresults are often difficult to predict andartefacts can appear. This needs anon-line pipeline for the data reduction.This would result in a more agile obser-vational procedure, enabling the ob-

server to follow through and concentrateon a winning trail.

Finally, it must be recalled that thereare different types of AO science tobe extracted from the AO observations:photometry, astrometry, mapping ofextended objects, mapping aroundbright sources, detection of faint com-panions to bright stars, polarisation, andspectra at low and high resolution. Theydiffer rather significantly by the datacharacteristics which they require, suchas angular resolution, Strehl ratio, ac-curacy in flat-fielding, extent of the corect-ed field, and S/N and, therefore, callfor different observing strategies.

Figure 1: Image of the laser plume in themesospheric sodium layer and the top of theRayleigh scattering cone. In this experiment(August 1998) the laser beam emitted by theAO+LGS ALFA system on the 3.5-m telescopeat Calar Alto is observed from the side with the2.2-m telescope located at 343 m from the 3.5-m telescope. The lines across the figure arestar trails. Abscisae: apparent width of thebeam in arcsec. Ordinates: left altitude in kmand right elevation from the bottom of the im-age in arcsec (F. Delplanque et al.).

2. Determination of the PSF andDeconvolution

Central to the AO data analysis is thedetermination of the PSF and the sub-sequent deconvolution. This question isbeing vigorously investigated and a num-ber of new developments and resultswere presented at the Conference. Theeffort is directed towards the precisemapping of the PSF envelope in radiusand azimuth. The method used initially,when there is no star in the field brightenough to provide an accurate PSF, wasto observe a nearby bright star before andafter the science object observations. Thishas proven to be unsatisfactory in mostcases because of significant changes inthe seeing conditions (and thus in the PSFprofile) occurring between calibrationand science exposures. New approach-es are being actively explored alone andin combination:

– The "synchronous PSF" method(developed by J.-P. Véran) makes use ofthe wavefront sensor, WFS, data to de-termine the on-axis PSF. The PSF so de-termined becomes a poorer match withfainter reference stars, increasing dis-tance from the reference star, and, ofcourse, worsening seeing.

The synchronous PSF method can, inprinciple, be extended to off-centre posi-tions but only if a SCIDAR is on site tomeasure the dependence on the profileof the turbulence with altitude.

– A method particularly suited to pho-tometry in crowded fields e.g. globularclusters, consists in starting from a giv-en (initially imprecisely known) PSF andusing the relatively isolated stars in thefield to improve the determination of thePSF. In the myopic deconvolution one canalso make use of a priori knowledge ofthe source’s shape (point-like, sharpedge, etc...).

– Myopic deconvolution (developed byJ. Christou, S. Jefferies and co-workers)makes use of multiple short exposures.The only unchanged pattern commonto all exposures is the science object.The method has been applied to re-solve images of bright close binariesobtained at 0.85 microns using the 941actuators AO system on the 3.5-m tele-scope of the Starfire Optical Range(SOR). The closest resolved separationis 0.067 arcsec for a difference in mag-nitude of 0.58.

This is a rich and complex topic forwhich only a schematic overview is giv-en here. Other aspects of post-process-ing were examined such as variousflavours of deconvolution procedures –non-linear deconvolution, Magain, Cour-bin & Sohy method, Richardson-Lucymethod, use of a different PSF at in-creasing radial distance from the refer-ence star and more.

Other issues were discussed whichwere related to data-acquisition strategy,flat-field, calibration, tethering, oversam-pling, non-common path aberration, andhow to remove the wings of the starimage, which is behind the coronograph.

ences will have to be understood, andtheir importance evaluated for differentscientific objectives and different hard-ware implementation of AO. In a firstphase, the exercise is applied to pho-tometry in crowded fields. Then, it will beextended to other types of data and as-trophysical projects: Fabry-Perot (FP)data, detection of companions, extendedemission, spectrography, polarisation,etc... opening a long-range process ofcontinuous improvement of AO dataanalysis.

Artefacts

Attention was drawn to the danger ofartefacts and spurious effects. A startlingillustration of this problem was present-ed in the form of images of the inner re-gion of galaxies harbouring an AGN tak-en with the Adaptive Optics Bonette(PUEO) of the CFHT. The artefacts forthese particular observations take theform of cross pattern structures. Theseare misleading as they look like thetorus+spiral arms structure expected, insome models, to be present in the im-mediate vicinity of AGN. The origin ofthese cross pattern structures is stillunclear but the moral of the story is thatextreme care and caution must be exer-cised in order to avoid this and other pit-falls. More efforts are needed to set upthe procedure that will remove these ef-fects.

The Superspeckles and the subtractionof images with identical PSF

There is exciting science, which re-quires the search, and discovery of veryfaint companions to bright stars i.e. browndwarfs and planets. The sensitivity of thesearch depends on the S/N in the enve-lope of the AO corrected star image. (Thefigure of merit here is the magnitude dif-

ference, ∆m, between the star and thecompanion.) The speckles produce acorrelated noise whose variance domi-nates the photon noise. To circumvent thisdifficulty, Racine, Walker and co-workershave devised the method of “DualImaging”: A beam-splitter produces twoimages of the field on the detector and anarrow-band filter centred on a spectralband which has a different strength in thecompanion and the star, is inserted in oneof the beams. The two images haveidentical PSF and differ only by the con-trast between the companion and thebright star. By subtracting one image fromthe other, one obtains an image where theresidual correlated speckle noise is re-duced by several orders of magnitude cor-respondingly increasing the S/N of thecompanion. The performance expectedfrom “Dual Imaging” is to increase ∆m by3 magnitudes and ∆m could even reach22 at 0.5 arcsec! The method of “DualImaging” can be used in other projects re-quiring the subtraction of two images:line/continuum observations, polarisa-tion and speckle imaging.

Astrometry

In star-forming regions, dust ab-sorption strongly modifies the distrib-ution of the visible light as comparedto that of the IR light resulting in differ-ences between the maps in the twobands. Exactly aligning the two maps isoften problematic (for example, in the ab-sence of foreground stars) and in thesecases, the colour and temperature of thesources cannot be measured with confi-dence. A simple solution is to have thepossibility to observe simultaneously inboth bands through two cameras, with adichroic separating the visible and IRbeams. For example, on NAOS a visiblecamera could be installed in the “paral-lel path”.

33

Figure 2: Two longitudinal profiles of the laser plume obtained at 20 min-utes interval during the experiment of Fig.1. One can note the changesin the profile near the maximum as well as the appearance of transientlayers around 76 and 87 km (F. Delplanque et al.).

Is there a bestdeconvolutionprocedure?

Could one pro-cedure be betteradapted to a givenastrophysical pro-gramme than an-other? In an ef-fort to answerthese questions,ESO has pro-posed to makeavailable a certainnumber of ADO-NIS data com-plete with theirPSF(s) and WFSdata to several in-terested groups.Each group willapply its algo-rithms and proce-dure to the samedata. The resultswill be compared.Significant differ-

3. AO Systems Coming intoOperation in the Near Futureon Very Large Telescopes

3.1 Systems commissioned nextyear, 1999, on 6- to 10-mtelescopes

– The University of Hawaii AO systemHokupa’a (Curvature sensor and 36 ac-tuators) will be set on the Gemini North8.2-m telescope.

– Japanese 8.2-m Subaru telescope.(Curvature sensor and 36 actuators).

– Keck 10-m telescope II (Shack-Hartmann sensor with eventually 349 ac-tuators) + laser guide star system, LGS.

– on Mt Hopkins, the “MMT” now retro-fitted with a 6.5-m single mirror (Adaptivesecondary with 330 actuators).

– It is recalled that the SOR 941 actu-ators system has started operation on the3.5-m telescope although not primarily forastronomical observations.

Furthermore, two spectrographs wherethe entrance slit is in the AO-corrected fo-cal plane have just started operation: theGraF Spectrograph on ADONIS at theESO 3.6-m telescope and the OASIS in-tegral field spectrograph for PUEO on theCFHT 3.6-m.

3.2 Systems commissioned in2000+ on 6- to 10-mtelescopes

In the year 2000 and shortly after, oth-er very large telescopes will be equippedwith AO systems, with LGS system inmost cases:

- on the VLT, NAOS+CONICA onUT1, SINFONI on UT1, CRIRES on UT4,MACAO for VLTI.

- on other very large telescopes:Keck I, the 2 LBT-8.4-m telescopes, the

8.2-m telescopes Gemini I and II, and theMagellan 6.5-m.

3.3 Ground-based Adaptive Opticsand Space Telescopes

Around the year 2007, when NGST (di-ameter about 8 m, wavelength range 0.8to 10 µ) is launched, about ten 6.5-m to10-m telescopes equipped with AO imag-ing and/or spectroscopy will have carriedout observations for 6–8 years, each witha larger collecting area and higher angularresolution than HST. There is currently anactive debate to define HST’s uniquenessspace in the era of AO systems on 8-mtelescopes. The result will be an impor-tant input in the definition of HST new in-struments and programmes, wide-fielddeep survey in J for example, and ofcourse UV and optical imaging andspectroscopy. (For more information onthis debate one can contact Jim Beleticat jbeletic@eso.org and Bob Fosbury atrfosbury@eso.org.)

Each of the two techniques, space-based observations and ground-basedAO observations, has its area of best per-formance. Complementarity is very fruit-ful (example in 5.1) – and some overlapis beneficial as a check.

The arguments of the debate in gen-eral, and also for the particular case ofGround-based AO versus NICMOS/HSTare well known: With equal diametertelescopes and equal integration time,broad-band space observations go deep-er than ground-based AO systems in theinfrared bands where the sky backgrounddominates. Other advantages of spaceare stable and uniform PSF, complete skycoverage, continuous wavelength cov-erage between 1 and 10 microns and be-yond and possibility to have the entiretelescope cooled. The advantages of

ground-based telescopes AO systems arethe diversity of instruments (such as nar-row-band filters, FP, spectrographs witha variety of resolutions), the easy upgradeof CCD, IR arrays and WFS, and the pos-sibility to carry out large programmes andextensive surveys since a number oflarge-diameter ground-based telescopesare being equipped with AO systems. Skycoverage is limited to less than 50 percent, the exact value depending on theStrehl ratio one could accept and on thegalactic latitude. In a few years, LGS willbring routinely nearly complete sky cov-erage and good PSFs.

Ground-based telescopes in the 2-mdiameter range could also be equippedwith AO systems getting close to the dif-fraction limit in the visible. The large num-ber of medium-size telescopes will makeit possible to carry out large scale pro-grammes and surveys.

4. Other Important Issues

4.1 Technologies – Laser GuideStar Systems

A number of technological advancesand experimental results were present-ed on deformable mirrors, adaptive sec-ondaries and laser guide-star systems(LGS). The first astrophysical resultsever obtained in Europe with an AO+LGSsystem were presented (Section 5.2).Monitoring of the column density and ef-fective altitude of the sodium mesosphericlayer is planned or under way at all ob-servatories building a LGS system andfirst results were presented. A large partof the LGS studies in Europe are co-ordinated through a network funded by theEuropean Union within its “Training andMobility for Researchers” (TMR) Pro-gramme. An example of such study is theobservation of the laser beam of the ALFAsystem at Calar Alto. In this experiment,the beam is launched from the 3.5-m tele-scope towards the zenith and is ob-served from the side with the 2.2-m tele-scope located 343 m from the 3.5-m. Thebeam excites the mesospheric sodiumlayer of about 10 km thickness located atan average altitude of 80 km. From theside the excited region is seen as a bright“plume”. The properties of this plume andtheir time variability are the object of thestudy. Some of the results obtained inAugust 1998 can be seen in Figures 1 and2 (F. Delplanque et al.). The 5-min. ex-posure on Figure 1 shows the plume, andbelow and above it the trail caused byRayleigh scattering, the intensity of thetrail reaching 1% of the maximum inten-sity of the plume at 7 arcmin below theplume centroid. The two curves onFigure 2 are the longitudinal (i.e. alongthe laser beam direction) profiles of theplume measured 20 minutes apart. In thistime interval, there are already noticeablevariations of the longitudinal profile and,moreover, one can note the appearanceof transient layers below and above themaximum intensity of the plume, here at76 and 85 km. These changes in shape

34

Figure 3: The jet of the pre-main sequence binary RW Aur imaged with PUEO at the CFHT, inthe optical emission lines [SII]6717,6734. The jet is associated with component A, marked bya cross. Resolution (FWHM) is 0.15″ and dynamic range 10 5, 3σ (F. Ménard et al.)

and in effective altitude have conse-quences on the wave-front sensor oper-ation. The shape modification increasesthe noise on the centroid measurementsin a different way on each sub-apertureand requires an adapted modal control.The altitude variations require that acontinuous measurement of the sodiumspot effective altitude be made in orderto correct for the induced defocus.Happily, these effects change relativelyslowly allowing for efficient correction. Thehindering effects of the light pollution forthe other telescopes on the site, and pos-sible solutions are also investigated.

4.2 Flexible Scheduling

The performance of an AO system,the Strehl ratio in particular, improves

35

FFigure 4: Comparison of the structure of theRW Aur A jet in the [SII] line and the [OI]6300line. The line intensity ratio varies along the jet;the [OI] emission decreases very rapidly alongthe jet and the first [OI] knot has no counter-part in [SII] (C. Dougados et al. quoted inMénard et al.).

Figure 5: (a) The polarisation vectors of the en-velope surrounding the double T Tauri systemUY Aur superposed on an image in J (Hokupa’aon the CFHT 3.6-m, November 1997). Note themaximum in polarisation at PA 130 and –40 de-grees, the centrosymmetric pattern in polari-sation angle and the high front/back contrastof the background image in J. (b) For comparison, a Mie scattering model ofa flattened dust disk with a power law (–4.7)size distribution of dust grain which best fits thefront/back ratio of intensity. Both figures fromD. Potter et al. FH(a) (b)

strongly with improving seeing. For bestreturn of money and time spent on build-ing the AO system and the telescope,the AO system should be used in excel-lent atmospheric conditions. FlexibleScheduling and Queue Scheduling aim atassigning the best conditions to the in-struments and programmes that needthem the most. This is a complicatedprocess, and requires to have on handthe statistical properties of the atmos-phere (r0, τ0, θ0) at the site on time scalesranging from hours to months. The ad-vantage of splitting nights with other in-struments will have to be clearlydemonstrated, taking into account cali-bration time and overhead time to passfrom the AO instrument to another andvice-versa. A recommended approach forearly use of the next generation of AOinstruments is to use it through entire(selected) nights, and execute the bestAO programmes suited to the currentconditions.

5. Astrophysical Results of AOObservations

10 papers on Circumstellar disks andjets, 8 on binaries and stars with faint com-panions, 7 on galaxies and quasars, and

36

2 on the Solar System were presented atthe Conference.

5.1 Circumstellar disks, Jets,and Stars

Star formation and Planet formation

AO observing in the near IR is partic-ularly well suited to this phenomenonwhich occurs in dust rich regions. The fi-nal mass of a star results from the com-bined effects of accretion through the diskand mass loss through jets and out-flows. Detailed maps, polarisation andspectroscopy provide clues to the fun-damental processes of star formation andearly evolution, and disk and jet forma-tion and evolution. The best laboratoriesin which to observe these phenomena arethe Classical T Tauri stars because, inthese systems, part of the dust has beencleared out making the base of the jets(and the disk) directly observable.Classical T Tauri are thus favourite tar-gets to investigate jet formation. Asan example, Figures 3 and 4 show the jetassociated with component A of the bi-nary Classical T Tauri star RW Aur,mapped in the optical range with PUEOat the CFHT (F. Ménard et al.). The fact

that the jet is very straight is an importantfinding about the location of the jet for-mation. Tidal torques being expected todisrupt or warp the disk, the jet’s straight-ness suggests that the jet originatesfrom the very central part of the disk whichremains unperturbed by the companionstar.

Only a handful of circumstellar disksaround T Tauri stars have been resolvedin their scattered light. There is very lim-ited information on the composition andsize distribution of the dust grains in thesedisks and on density inhomogeneities.This information, however, is crucialto the understanding of how planetsform. The combination of infrared imag-ing and polarisation is a powerful tool togather this information as is illustrated inthe images of the circumbinary diskaround UY Aurigae (Figure 5a, b from D.Potter et al.) obtained with Hokupa’a onthe 3.6-m CFHT in November 1997.Figure 5a shows the high-contrast, high-resolution polarimetry map with the po-larisation vectors overlaid on the de-convolved grey-scale image in J. The po-larisation vectors show a centrosymmet-ric pattern in angle and intensity withpeaks in polarisation intensity at positionangle of 130 and –40 degrees. This pat-tern is entirely consistent (Figure 5b) withthe extended IR surface brightness ob-served around UY Aur being a flatteneddust disk with a large number of smallgrains. The front/back ratio of intensityconstrains the power law size distributionof the grains. The alternative model of aspherical envelope of dust appears un-likely.

Results obtained on the young massivestar Her 36 also illustrate the power of AOobservations and give an example of thecomplementarily of AO, HST and radioobservations (Figures 6 and 7 from B.Stecklum et al.). Her 36, spectral typeO7V, is responsible for the ionisation ofthe “Hourglass” Nebula. Early photo-metric and polarimetric investigationsyielded clear evidence that it showedunexpected peculiarities in polarisationand a strong infrared excess. High an-gular resolution in the IR, first by lunar oc-cultation, then with ADONIS, revealedthe intense IR source to be not Her36 itself but a separate object locat-ed 0.3″, or 500 AU from Her 36 in pro-jection (Figure 6). This clarifies the natureof Her 36 as a normal young massive star,but the nature of the intense IR source re-mains a mystery. HST detected thesource in Hα and the VLA detected it asa weak point source at 2 cm (Figure 7).The search and non-detection in maseremission makes it unlikely that the IRsource is a deeply embedded massiveyoung stellar system. Could it be disk de-bris left from the formation of Her 36, ora proplyd object, or an evaporatinggaseous globule?

Main Sequence Stars

Circumstellar disks are also observedaround MS stars. The best known ex-

Figure 6: The young massive star Her 36 and the mysterious intense infrared source form thedouble object in the lower half of the image. The IR source is at 0.3", or 500 AU in projection,from Her 36; its nature is still unknown. Observations with ADONIS in L (B. Stecklum et al.).

ample is β Pictoris. By coronographywith ADONIS, the disk has been mappedto within 20 AU from the star, that isinside the planet orbits in our SolarSystem!

Several papers were also given on thefrequency of occurrence of binary starsand its dependence on environment andage, and on the search for brown dwarfsand planetary companions. The imageanalysis strategy can take advantage ofthe a priori information that the sources

in the field are point-like. This is valid alsofor that uniquely important target, theGalactic Centre.

5.2 Extragalactic Studies

Results have been reported on gravi-tational lenses, intervening galaxies, andquasar companions. Attempts have beenmade at finding signs of accretion in thevicinity of AGN or at identifying structureswhich could be the molecular torus ex-

pected to be present around AGNs.Maps of luminous IR galaxies with un-precedented resolutions were presented.The compactness of the IR central sourcehelps to distinguish between the presenceof a dust embedded quasar or of a com-pact bright star-forming region.

At the Conference were presentedthe first images of galaxies ever ob-tained in Europe with an AO+LGS sys-tem (R. Davies et al.). They are part of aprogramme to observe the light distribu-tion in the core of galaxies in clusters. Inthe clusters studied, there appears to bea correlation between the size of the coreat the galaxies’ centre with the radial dis-tance of the galaxies from the cluster’scentre and their HI content – the core be-ing more compact in more centrally lo-cated galaxies. Davies et al. suggest thatthe crossing of the central part of the clus-ter by a galaxy depletes the galaxy of itsHI (a long-known effect) and stimulatesstar formation in the gas left in the galaxycore, gas which is likely to be subjectedto strong tidal effects.

5.3 Solar System

In the age of space probes, there re-main important Solar System observa-tions, which are most successfully carriedout with AO on large telescopes. AO candetect and monitor faint, variable or tran-sient objects/structures on the planetarysurfaces, rings, and satellites. Further-more, AO yields the most detailed mapsof solar-system objects not targeted by thespace probes. One of the impressive plan-etary observations presented in Sont-hofen is shown in Figure 8 (F. Roddier etal.). The three images of Neptune weretaken 45 min apart on July 6, 1998. Theywere taken through a narrow-band filtercentred inside a methane absorptionband at 1.72 µ, with Hokupa’a on theCFHT. The angular diameter of Neptunewas 2.36″. High-altitude clouds are seenhere against a dark methane band at-mospheric absorption and appear bright.During the Voyager encounter, most ofNeptune’s cloud activity was in the South

37

Figure 7: HST/WFPC2 Hα image of Her 36 after PSF subtraction. The intense IR source is de-tected as a weak point-like source at 2 cm with the VLA (contour lines) but is not detected inmaser emission making it unlikely that it is a deeply embedded young stellar object (B. Stecklumet al.).

Figure 8: Cloud activity on Neptune. Images taken 45 min apart with Hokupa’a on the CFHT on July 6, 1998. High-altitude clouds appear brightagainst a dark methane band atmospheric absorption. Note the rapid planet rotation. The angular diameter of Neptune was 2.36 ″ (F. Roddieret al.).

hemisphere. It then moved to the northwhere it peaked around 1994. In July1998, as can be seen on the Figure, thecloud activity had grown again in thesouth. The causes of these changes areunknown.

6. Conclusions

This first Conference centred onAstrophysics with AO was a success, andthe common wish at the end was tohave such a Conference on “Astronomy

with Adaptive Optics” every 1.5 years.The Adaptive Optics systems expect-

ed to be routinely operating in the nearfuture are growing in number. The ob-servers are getting more and more fa-miliar and proficient with their “new tool”and define new strategies to observe ef-ficiently. We can expect a large growthof striking results in the coming years aswell as a continuation of the technologymaturation process. This continues to mo-tivate ESO’s effort to equip the VLT withseveral AO systems in the near future.

Moreover, these efforts pave the way tothe road to the Giant Optical Telescopesof the next century.The paper version of the Proceedings canbe obtained by writing to Christina Stofferat ESO, cstoffer@eso.org. On-line ver-sion of the papers is available athttp://www.eso.org/gen-fac/meetings/eso-osa98

38

mhulrich @eso.org

Award for the ESO VLT Project Naturally, ‘First Light’ for the first of the

VLT Unit Telescopes has not gone un-noticed by the astronomical community.However, also newspapers, magazines,radio and TV have reported widely aboutthis new facility and about the first, fas-cinating pictures that were obtained withUT1. Another sign of the world-wide at-tention that the VLT has caught was thefact that the well-known US magazinePopular Science chose the project for oneof the coveted “The Best of What’s NewAwards” for 1998, given to the 100 mostimportant technology developments in thecourse of the year. The construction of theVLT was recognised as an “outstandingachievement” within the “Aviation andSpace” category.

The Award was formally handed overto ESO, represented by Prof. MassimoTarenghi, during a luncheon at the Tavernon the Green in New York on November13. On the same occasion, about 500 in-vited guests from industry, the media and

government agencies had the opportunityto inform themselves about the individualprize winners through project and prod-uct presentations by means of small ex-hibitions. Here, ESO found itself in thecompany of a great variety of organisa-tions and enterprises working in differenthigh-tech areas, ranging from NASA(International Space Station and theMars rover “Sojourner”), Airbus Industrie(A-340-500) and various experimental air-craft manufacturers, through IBM (copper-based computer chips), Apple Computer(iMac), and Iridium Satellite telephonesto the latest products in medicine andmedical research.

Given the audience and the extreme-ly high technological level of the projectsand innovations presented, this was un-doubtedly a good occasion for ESO to in-form US media and the technologically in-terested public about the VLT as majorEuropean science project.

C. MADSEN

M. Tarenghi, posing for the official photo at the ESO Stand.

Popular Science editor Cecilia Wessner speaking at the award ceremony.

39

A N N O U N C E M E N T S

ESO Director General to Become President of AUITEXT OF THE JOINT ESO-AUI PRESS RELEASE OF 3 NOVEMBER 1998 (E XCERPTS)

The appointment of Professor Riccardo Giacconi, Director General of the European Southern Observatory (ESO) since January 1, 1993,to the Presidency of Associated Universities, Inc. (AUI) in the USA, has been jointly announced by Professor Paul C. Martin, Chair of AUI’sBoard of Trustees and Mr. Henrik Grage, President of the ESO Council. Professor Giacconi will assume this new position at the end of histerm at ESO as of July 1, 1999.

AUI is a not-for-profit science management corporation that operates the National Radio Astronomy Observatory (NRAO) under a CooperativeAgreement with the National Science Foundation (NSF). Corporate headquarters are located in Washington, D.C. The President is its chiefexecutive officer.

Mr. Grage, President of the ESO Council, expressed the gratitude of the ESO Community for the leadership provided by Prof. Giacconiduring these crucial years of development of the organisation and its La Silla and Paranal Observatories. In particular, the splendid achieve-ments on the Very Large Telescope (VLT) are a tribute to the ESO staff and to his management and guidance. VLT is currently the largestsingle project in ground-based astronomy. It has met or exceeded all performance requirements while being built on time and within bud-get.

When reached for comment, Professor Giacconi pointed out: “I have enjoyed enormously the time I have spent here at ESO and I con-sider it one of the high points of my career. I feel confident that I am leaving ESO in very good condition. The fine performance of the entirestaff has succeeded in bringing the organisation to an outstanding position in ground-based astronomy in the world. The prospects for thefuture are equally brilliant.

I will be happy and proud to assume the Presidency of Associated Universities, Inc. starting next summer. For more than fifty years, AUIhas, in collaboration with universities and the national and international scientific community, overseen and managed national facilities whichhave made possible a wealth of important discoveries in physics, astronomy, and many other areas of science and technology. In the 21stcentury, new challenges and opportunities to serve the community await AUI.”

“We are thrilled that Professor Giacconi has decided to take this position", said Professor Paul Martin, Chairman of the Board of AUI. "Itis hard to imagine anyone better qualifıed to lead an organisation committed to managing facilities performing frontier science. His visionand foresight have been at the heart of pioneering projects including the Einstein Observatory, the Space Telescope, and the VLT. He is anextraordinary scientist and an outstanding manager whose accomplishments and values have earned him worldwide respect and admira-tion.”

F I R S T A N N O U N C E M E N T

ESO Workshop on

Black Holes in Binaries and Galactic Nuclei –Their Diagnostics, Demography, and Formation

6–8 September 1999ESO Headquarters, Garching bei München, Germany

The aim of this workshop is to bring together astrophysicists working on the rather separate fields of stellar mass and supermassive blackholes, with emphasis on the formation, population, physics, and environments of black holes. An overview will be given of the observation-al evidence for the existence of black holes in binaries and galactic nuclei. Related topics like the late stages of (binary) evolution and galaxyformation will be addressed.

Topics to be covered include:• Black-hole diagnosticsBinaries (X-ray sources, neutron stars vs. black holes)Centres of galaxies, including our own (AGN phenomenon, stellar and gas dynamics signatures)Jets, disks and accretion tori; Variability and outbursts

• Properties of black-hole populations Stellar-mass black holes; Supermassive black holes

• Black-hole formation Progenitors of stellar-mass black holes: Supernovae/Hypernovae Models for the formation of supermassive black holes

Scientific Organising Committee:R. Bacon (France), R. Bender (Germany), R.D. Blandford (USA), P.T. de Zeeuw (Netherlands), L. Kaper (Netherlands),G. Monnet (ESO), M.J. Rees (United Kingdom), A. Renzini (ESO), D.O. Richstone (USA), R. Sunyaev (Germany), E.P.J. van den Heuvel (Netherlands, Chair)

Local Organising Committee:L. Kaper (Netherlands, Chair), C. Stoffer (ESO), P. Woudt (ESO)European Southern ObservatoryKarl-Schwarzschild-Str. 2D-85748 Garching bei München, GermanyTel: +49-89-320060; FAX: +49-89-32006480Email: bh99@eso.orgSee also: http://www.eso.org/bh99/

40

VLT OPENING SYMPOSIUMAntofagasta, Chile, 1–4 March 1999

To mark the beginning of the VLT era, the European Southern Observatory is organising a VLT Opening Symposium that will take placein Antofagasta on 1–4 March 1999, just before the start of regular observations with the VLT.

The Symposium will occupy four full days and consist of plenary sessions (invited talks only) on “Science in the VLT Era and Beyond” andthree parallel Workshops (invited and contributed talks) on “Clusters of Galaxies at High Redshift”, “Star-way to the Universe” and “FromExtrasolar Planets to Brown Dwarfs”.

During the Symposium, participants will have the opportunity to see the VLT in operation. Special visits to the Paranal Observatory willbe organised each afternoon and early evening. After the Symposium, there will be an Official Inauguration Ceremony at Paranal on March5. The Ceremony is intended largely for official guests from the member states and Chile and will be by invitation only.

Plenary Sessions: SCIENCE IN THE VLT ERA AND BEYOND Scientific Organising Committee: D. Alloin, I. Appenzeller, J. Bergeron (co-Chairperson), C. de Bergh, B. Fort, R. Kudritzki, M. Mayor,

G. Miley, P. Nissen, H. Quintana, A. Renzini (co-Chairperson), R. Sancisi

List of topics:• Cosmology: where do we stand? – S. White • The centres of nearby galaxies – R. Bender• The most distant galaxies – M. Pettini • Nucleosynthesis in the young Universe – B. Gustafsson• Galaxy evolution and nuclear activity – G. Miley • New frontiers in the solar system – A. Barucci• Probing active galaxies in the infrared – A. Moorwood • Science with the Large Southern Array – K. Menten• Supernovae at high redshift – R. Kirshner* • Synergy between the VLT and the NGST – H. Stockman• The cosmic history of star formation – P. Madau * Tentative speaker

The last plenary session will include a presentation of the highlights of the three workshops:

• Clusters of Galaxies at High Redshift • From Extrasolar Planets to Brown Dwarfs• Star-way to the Universe

Parallel Workshop 1: CLUSTERS OF GALAXIES AT HIGH REDSHIFTSOC: H. Böhringer, A. Cavaliere, B. Fort (co-Chairperson), P. Katgert, P. Rosati (co-Chairperson), P. Schneider, M. Tarenghi

Preliminary list of topics:• Clusters at z >1 (optical, IR) • Cluster masses and dynamics• X-ray clusters at high z • Weak lensing• SZ clusters and surveys • Probing very distant galaxies with lensing clusters• LSS at high-z and cluster formation (theory) • ICM physics and evolution• LSS at very high redshift (observations) • Fundamental plane at high z• Clustering from absorption systems • ESO Imaging Survey• Clustering around high-z radio galaxies • 2dF Survey• Galaxy populations and galaxy evolution in high-z clusters

Parallel Workshop 2: STAR-WAY TO THE UNIVERSESOC: I. Appenzeller, C. Chiosi, R. Kudritzki (co-Chairperson), J.-C. Mermilliod, P. Nissen, L. Pasquini (co-Chairperson), M. Spite

Preliminary list of topics: • The VLT-UT2: a unique machine for stellar studies• Distance Indicators: Cepheids. Improvement of Cepheids as distance indicators. Eclipsing binaries as precise distance indicators. Eclipsingbinaries. Supergiants. Extragalactic planetary nebulae and what to do with them. Novae and GC systems• Age Indicators: Age of globular clusters. MC’s clusters. Globular clusters in the Galactic bulge. Bulge of our Galaxy. Other galaxies. Starforming regions. WR stars in star forming regions. Age indicators in solar-type and low-mass stars • Abundance Indicators: Massive stars and their role in spectra and abundances of high redshift galaxies. Massive stars: observation.Intermediate/low mass in Sagittarius. Intermediate/low mass in MC’s. The deep thick disk as seen by UVES. Intermediate/low mass in theBulge. Stars versus HII regions: extragalactic HII regions. Stars versus HII regions: extragalactic stars

Parallel Workshop 3: THE SEARCH FOR EXTRASOLAR PLANETS WITH THE VLT/VLTISOC: F. Allard, P. Artymowicz, M. Mayor (co-Chairperson), F. Paresce (co-Chairperson), M.-T. Ruiz, P. Sackett

Preliminary list of topics: Formation and Evolution: Reviews• Theory: Diversity of planetary systems: formation scenarios. Brown dwarfs formation and evolution theories. Unsolved problems and fu-ture directions in theory. Solar nebulae and replenished disks with planets. Pulsar planet theories. Evolution of brown dwarfs and giant plan-ets: contribution to the Galactic mass budget. The stellar and substellar IMF: theory• Observations: Reviews: Protoplanetary disks. Replenished dust disks. The low-mass stellar IMF. The substellar IMF. The spectra of giantplanets and brown dwarfs

Techniques• Radial Velocity Searches: The large planet survey with CORALIE at La Silla. The Keck radial velocity search for extrasolar planets andbrown dwarfs. Visible or infrared VLT high-resolution spectrographs (UVES and CRIRES). Towards 1 m/s with the 3.6-m telescope at LaSilla• Astrometric Detections: From PTI to KECK Interferometer. Astrometry with the VLTI. Astrometry with the NPOI. Astrometry from space:GAIA. Astrometry from space: SIM• Planetary Microlensing and Transits: Planetary microlensing: present status and long term goals. Planetary transits• Direct Detections: Brown dwarfs with DENIS. Brown dwarfs with 2MASS. Brown dwarfs in open clusters. Interferometric detections of gi-ant planets and brown dwarfs with the VLTI. Adaptive optics from the ground or from space. NAOS-CONICA. VIZIR. IRSI

Local Organising Committee: D. Alloin, D. HofstadtClosing date for registration is December 31, 1998. More details and a registration form can be found at: URL http://www.eso.org/gen-

fac/meetings/vltos99/ or contact vltos99@eso.org for further information.

41

1302. G.F. Lewis et al.: Submillimeter Observations of the UltraluminousBAL Quasar APM 08279+5255. ApJ Letters.

1303. R. A. Ibata and G. F. Lewis: Optimal Proper MotionMeasurements with the Wide Field and Planetary Camera.AJ.

1304. P. Andreani et al.: The Enhancement and Decrement of theSunyaev-Zeldovich Effect Towards the ROSAT Cluster RXJ0658-5557. ApJ.

1305. E.M. Corsini et al.: Dark Matter in Early-Type Spiral Galaxies:The Case of NGC 2179 and of NGC 2775. A&A.

1306. M. Kissler-Patig et al.: Towards an Understanding of theGlobular Cluster Over-Abundance Around the Central GiantElliptical NGC 1399. A&A.

1307. L. Pulone, G. De Marchi and F. Paresce: The Mass Function ofM4 from Near IR and Optical HST Observations. A&A.

1308. G. De Marchi, B. Leibundgut, F. Paresce, and L. Pulone: A DeepOptical Luminosity Function of NGC 6712 with the VLT: Evidencefor Severe Tidal Disruption. A&A.

1309. A. Pasquali, A. Nota, M. Clampin: Spatially Resolved NebulaeAround the Ofpe/WN Stars S61 and Be381. A&A.

1310. F. Primas, D.K. Duncan, R.C. Peterson, J.A. Thorburn: A NewSet of HST Boron Observations. I. Testing Light Elements StellarDepletion. A&A.

1311. L. Kaper, H.F. Henrichs, J.S. Nichols, and J.H. Telting: Long-and Short-Term Variability in O-Star Winds. II. QuantitativeAnalysis of DAC Behaviour. A&A.

1312. F. Comerón, G.H. Rieke, R. Neuhäuser: Faint Members of theChamaeleon I Cloud. A&A.

List of Scientific Preprints(September–December 1998)

1293. A.R. Tieftrunk, R.A. Gaume, T.L. Wilson: High-ResolutionImaging of NH3 Inversion Lines Toward W3 Main. A&A.

1294. F. Primas, D.K. Duncan, J.A. Thorburn: The Remarkable Boron-Depleted, Lithium-Normal Population II Star HD 160617. ApJLetters.

1295. D.R. Silva, G.D. Bothun: The Ages of Disturbed Field EllipticalGalaxies: II. Nuclear Properties. AJ.

1296. S. Brillant, G. Mathys, C. Stehlé: Hydrogen Line Formation inDense Magnetized Plasmas. A&A.

1297. C.R. Cowley and G. Mathys: Line Identifications and PreliminaryAbundances from the Red Spectrum of HD 101065 (Przybylski’sStar). A&A.

1298. S. Hubrig, F. Castelli and G. Mathys: Isotopic Composition ofHg and Pt in 5 Slowly Rotating HgMn Stars. A&A.

1299. G. Mathys: Direct Observational Evidence for Magnetic Fieldsin Hot Stars. Variable and non-spherical stellar winds in lumi-nous hot stars, Proc. IAU Coll. B. Wolf, A. Fullerton, O. Stahl(eds.).

1300. B. Leibundgut and J.G. Robertson: Emission Within a DampedLyman α Absorption Trough: The Complex Sight Line TowardsQ2059–360. MNRAS.

1301. R.A. Ibata, G.F. Lewis and J.-P. Beaulieu: Re-examination ofthe Possible Tidal Stream in Front of the LMC. ApJ Letters.

ESO Workshop ProceedingsStill Available

Most ESO Conference and Workshop Proceedings are still avail-able and may be ordered at the European Southern Observatory.Some of the more recent ones are listed below.

No. Title Price

50 Handling & Archiving Data from Ground-based Telescopes. Trieste, Italy, April 21–23, 1993. M. Albrecht & F. Pasian (eds.) DM 35.–

51 Third CTIO/ESO Workshop on “The Local Group: Comparative and Global Properties”. La Serena, Chile, 25–28 January 1994. M. Albrecht and F. Pasian (eds.) DM 35.–

52 European SL-9 Jupiter Workshop. February 13–15, 1995, Garching, Germany. R. West andH. Böhnhardt (eds.) DM 80.–

53 ESO/ST-ECF Workshop on “Calibrating and understanding HST and ESO instruments”, Garching, Germany. P. Benvenuti (ed.) DM 60.–

54 Topical Meeting on “Adaptive Optics”, October 2–6, 1995, Garching, Germany. M. Cullum (ed.) DM 80.–

55 NICMOS and the VLT. A New Era of High Reso- lution Near Infrared Imaging and Spectroscopy. Pula, Sardinia, Italy, May 26–27, 1998 DM 20.–

PERSONNEL MOVEMENTSInternational Staff

(1 October – 31 December 1998)

ARRIVALS

Europe

CLOSE, Laird (CDN), Adaptive Optics Instrument ScientistDAY, Philippa (GB), Technical Documentation ArchivistSKOLE, Steen (DK), Software EngineerALVES, Joao (P), Fellow CONTINI, Thierry (F), FellowROMANIELLO, Martino (I), FellowSUCHAR, Dieter (D), Archive Operations Systems

AdministratorZAMPARELLI; Michele (I), Systems Engineering Group of DMDTORDO, Sébastien (F), CoopérantNAUMANN, Michael (D), WEB Administrator/MasterCAVADORE, Cyril (F), CCD Detector SpecialistPATIG, Markus (D), FellowFERRARI, Marc (F), AssociateLEVEQUE, Samuel (F), Photonics PhysicistPRIETO, Almudena (E), User Support Astronomer

Chile

GARCIA AGUIAR, Martina (D), Mechanical EngineerCHEVASSUT, Jean-Louis (F), DENIS ProjectELS, Sebastian (D), StudentJOUGET, Benoit (F), CoopérantHÖÖG, Torbjörn (S), Mechanical/Electrical Engineer-Deputy

of the Head of DepartmentDELAHODDE, Catherine (F), StudentENDL, Michael (A), StudentMARCHIS, Franck (F), StudentSEKIGUCHI, Tomohiko (Japan), Student

DEPARTURES

Europe

HOOK, Isobel (GB), FellowVAN DE SPRENG, Jacob (NL), Project Control OfficerWOLTERS, Otto (NL), Accounting AssistantLEVEQUE, Samuel (F), Associate

DUDZIAK, Gregory (F), Student and AssociateDE MARCHI, Guido (I), FellowBERNARDI, Mariangela (I), StudentERBEN, Thomas (D), Student and EIS VisitorPULONE, Luigi (I), Associate DGDF

Chile

TIEFTRUNK, Achim (D), FellowCHIESA, Marco (D), Telescope Software EngineerBOSKER, Albert (NL), Administrative AssistantGONZALEZ, Jean-François (F), Fellow

Organisational MattersR. Giacconi: The Role of ESO in European

Astronomy 91, 1R. Giacconi: First Light of the VLT Unit

Telescope 1 92, 2The Final Steps Before “First Light” 92, 3R.West: VLT First Light and the Public 92, 4

Observing with the VLT

B. Leibundgut, G. de Marchi, A. Renzini:Science Verification of the VLT UnitTelescope 1 92, 5

Scientific Evaluation of VLT-UT1 Proposals.Observing Period 63 (April 1 – October 1,1999) 92, 9

P. Quinn, J. Breysacher, D. Silva: First Call forProposals 92, 9

R.G. Petrov, F. Malbet, A. Richichi, K.-H. Hof-mann: AMBER, the Near-lnfrared/Red VLTIFocal Instrument 92, 11

The VLT-UT1 Science Verification Team:Science Verification Observations on VLT-UT1 Completed 93, 1

M. Sarazin: Astroclimate During ScienceVerification 93, 2

M. Tarenghi, P. Gray, J. Spyromilio, R. Gil-mozzi: The First Steps of UT1 93, 4

Portuguese Minister of Science at Paranal 93,7

The Cost of the VLT 93, 8B. Koehler: ESO and AMOS Signed Contract

for the VLTI Auxiliary Telescopes 93, 11 B. Koehler, F. Koch: UT1 Passes “With

Honour” the First Severe Stability tests forVLTI 93, 11

Signing of Contract for the Delivery of the DelayLine of the VLTI 91, 25

Image from the VLT Science VerificationProgramme 93, 29

News from VLT Science Verification 94, 9

Telescopesand InstrumentationA. Moorwood, J.-G. Cuby and C. Lidman: SOFI

Sees First Light at the NTT 91, 9S. D’Odorico: The New Direct CCD Imaging

Camera at the NTT. SUSI2 gets the firstFIERA CCD controller and a mosaic of two2k × 4k, 15µ pixel, thinned, anti-reflectioncoated EEV chips 91, 14

J.G. Cuby and R. Gilmozzi: VIMOS and NIR-MOS: Status Report 91, 16

P.O. Lagage, Y. Rio, J.W. Pel, H. Toisma:VISIR at PDR 91, 17

D. Baade et al.: The Wide Field Imager for the2.2-m MPG/ESO Telescope: a Preview93, 13

I. Appenzeller, K. Fricke, W. Fürtig, W. Gässler,R. Häfner, R. Harke, H.-J. Hess, W. Hum-mel, P. Jürgens, R.-P. Kudritzki, K.-H. Man-tel, W. Meisl, B. Muschielok, H. Nicklas, G.Rupprecht, W. Seifert, O. Stahl, T. Szeifert,K. Tarantik: Successful Commissioning ofFORS1 – the First Optical Instrument onthe VLT 94, 1

A. Moorwood, J.-G. Cuby, P. Biereichel, J.Brynnel, B. Delabre, N. Devillard, A. VanDijsseldonk, G. Finger, H. Gemperlein, R.Gilmozzi, T. Herlin, G. Huster, J. Knudstrup,C. Lidman, J.-L. Lizon, H. Mehrgan, M.Meyer, G. Nicolini, M. Petr, J. Spyromilio,J. Stegmeier: ISAAC Sees First Lightat the VLT 94, 7

S. Stanghellini and A. Michel: Performance ofthe First Two Beryllium Secondary Mirrorsof the VLT 94, 10

NEWS FROM THE NTTG. Mathys: News from the NTT 91, 21G. Mathys: News from the NTT 92, 14

THE LA SILLA NEWS PAGE

S. Guisard: Image Quality of the 3.6-mTelescope (Part VlII). New Seeing Recordof 0.47″ 91, 24

R. Gredel and P. Leisy: EFOSC2 and VLTAutoguider Commissioned at the 3.6-mTelescope 91, 24

Ph. Gitton and L. Noethe: Tuning of theNTTAlignment 92, 15

M. Kürster: CES Very Long Camera Installed92, 18

J. Brewer and J. Andersen: Improving ImageQuality at the Danish 1.54-m Tele-scope 92, 18

O.R. Hainaut: News from the NTT 93, 16C. Lidman: SOFI Receives its First Users 93,

16M. Sterzik: 3.6-m Telescope Passes Major

Upgrade Milestones 93, 17The 2p2team: 2.2-m Telescope Upgrade

Started 93, 19O.R. Hainaut: News from the NTT 94, 11T. Augusteijn: 2.2-m Telescope Upgrade: a

Status Report 94, 12M. Kürster: Performance of CES 3.6-m Fibre

Link and Image Slicers 94, 12

VLT Data Flow OperationsNewsD. Silva: The NTT Service Observing Pro-

gramme: Period 60. Summary and LessonsLearned 92, 20

A. Moorwood: SOFI Infrared Images of the“NTT Deep Field” 92, 25

ESO and ST-ECF ArchiveNews

B. Pirenne, M. Albrecht, B. Leibundgut: ESOand ST-ECF Archive News 93, 20

M. Albrecht: The VLT Data Volume 93, 21B. Pirenne, M. Albrecht: Using DVD Technol-

ogy for Archiving Astronomical Data 93, 22M. Dolenski, A. Micol, B. Pirenne, M. Rosa:

How the Analysis of HST EngineeringTelemetry Supports the WFPC2 AssociationProject and Enhances FOS CalibrationAccuracy 93, 23

B. Pirenne, B. McLean, B. Lasker: ST-ECFParticipation in the GSC-II GenerationProject 93, 25

M. Dolenski, A. Micol, B. Pirenne: HST ArchiveServices Implemented in Java 93, 26

A. Micol, D. Durand, S. Gaudet, B. Pirenne:HST Archive News: On the Fly Recalibration(OTF) of NICMOS and STIS Data 93, 27

A. Micol, B. Pirenne: HST Archive News:WFPC2 Associations 93, 28

Observations with theUpgraded NTT

P. Bonifacio and P. Molaro: Upgraded NTTProvides Insights Into Cosmic BigBang 92, 26

R. Neuhäuser, H.-C. Thomas, F.M. Walter:Ground-Based Detection of the IsolatedNeutron Star RXJ185635-3754 at V = 25.7Mag with the Upgraded NTT 92, 27

I. Burud, F. Courbin, C. Lidman, G. Meylan, P.Magain, A.O. Jaunsen, J. Hjorth, R.Ostensen, M.l. Andersen, J.W. Clasen, R.Stabell, S. Refsdal: RXJ0911.4+0551: A

Complex Quadruply Imaged GravitationallyLensed QSO 92, 29

The Large Southern ArrayP.A. Shaver and R.S. Booth: The Large

Southern Array 91, 26A. Otárola, G. Delgado, R. Booth, V. Belitsky,

D. Urbain, S. Radford, D. Hofstadt, L.Nyman, P. Shaver, R. Hills: European SiteTesting at Chajnantor 94, 13

SEST Upgrades and Reports from SEST ObserversL.-Å. Nyman and A. Tieftrunk: SEST Upgrades

91, 28 F. Combes and T. Wiklind: Molecular Lines in

Absorption at High Redshift 91, 29P. Andreani: Using SEST to Probe the Geom-

etry of the Universe 91, 35 R. Chini and E. Krügel: Cold Dust in Galaxies

91, 37J. Lequeux: Carbon Monoxyde in the Magel-

lanic Clouds 91, 41C. Henkel, Y.-N. Chin, R. Wielebinski, R.

Mauersberger: Cool Gas in SouthernGalaxies 91, 45

Science with the VLT/VLTIL. Da Costa, E. Bertin, E. Deul, T. Erben, W.

Freudling, M.D. Guarnieri, I. Hook, R. Hook,R. Mendez, M. Nonino, L. Olsen, l. Prandoni,A. Renzini, S. Savaglio, M. Scodeggio, D.Silva, R. Slijkhuis, A. Wicenec, R. Wich-mann, C. Benoist: The ESO Imaging Sur-vey: Status Report and Preliminary Results91, 49

A. Renzini: Après EIS 91, 54M. Arnaboldi, M. Capaccioli, D. Mancini, P. Ra-

fanelli, R. Scaramella, G. Sedmak, G.P. Vet-tolani: VST: VLT Survey Telescope 93, 30

Reports from ObserversJ. Sanner, M. Geffert, H. Böhnhardt, A. Fiedler:

Astrometry of Comet 46P/Wirtanen at ESO:Preparation of ESA’s ROSETTA Mission92, 33

L. Binette, B. Joguet, J.C.L. Wang, G. MagrisC.: NTT Archives: the Lyman α Profile of theRadio Galaxy 1243+036 Revisited 92, 37

A.R. Tieftrunk, S.T. Megeath: Star FormationToward the “Quiescent” Core NGC 6334I(N) 93, 36

M. Rubio, G. Garay’ R. Probst: Molecular Gasin 30 Doradus 93, 38

H. Böhringer, L. Guzzo, C.A. Collins, D.M. Neu-mann, S. Schindler, P. Schuecker, R. Crud-dace, S. Degrandi, G. Chincarini, A.C.Edge, H.T. Macgillivray, P. Shaver, G.Vettolani, W. Voges: Probing the CosmicLarge-Scale Structure with the REFLEXCluster Survey: Profile of an ESO KeyProgramme 94, 21

R.P. Mignani, S. Mereghetti, C. Gouiffes andP.A. Caraveo: Timing, Spectroscopy andMulticolour Imaging of the Candidate Op-tical Counterpart of PSR B1509–58 94, 25

F. Comerón, R. Neuhäuser, A.A. Kaas: TheFirst X-ray Emitting Brown Dwarf 94, 28

Call for Ideas for Future Public ImagingSurveys 94, 31

Other Astronomical NewsPhotos from Science Writers’ Symposium 92,

39B. Nordström: A Fresh Look at the Future: "La

Silla 2000++" 93, 42

42

MESSENGER INDEX 1998 (Nos . 91–94)

S U B J E C T I N D E X

M.-P. Véron, G. Meylan: 6th ESO/OHP Sum-mer School in Astrophysical Observations93, 43

C. Madsen, R. West: Sea & Space – A Suc-cessful Educational Project for Europe’sSecondary Schools 93, 44

M.-H. Ulrich: Observations with AdaptiveOptics: Getting There 94, 32

C. Madsen: Award for the ESO VLT Project94, 38

AnnouncementsESO Studentship Programme 91, 58Second Announcement of the ESO/ST-ECF

Workshop on NICMOS and the VLT: A NewEra of High-Resolution Near-lnfrared Imag-ing and Spectroscopy 91, 58

List of ESO Scientific Preprints 91, 58 The 34th Liège International Astrophysics

Colloquium “The Next-Generation SpaceTelescope – Science Drivers and Techno-logical Challenges” 91, 59

Personnel Movements 91, 59 Joint Committee Between ESO and the

Government of Chile for the Developmentof Astronomy in Chile 92, 40

L. da Costa and A. Renzini: ESO ImagingSurvey: Update on ElS-deep and the HubbleDeep Field South 92, 40

ESO Fellowship Programme 1999 92, 41First Announcement of an “ESO Conference

on Chemical Evolution from Zero to HighRedshift” 92, 42

TheAstronomical Almanac to be Revised 92, 42List of Scientific Preprints (March-May 1998)

92, 42Preliminary Announcement of an “ESO Work-

shop on Minor Bodies in the Outer SolarSystem” 92, 43

Personnel Movements (1 April – 30 June 1998)92, 43

List of New ESO Publications 93, 46Pierre Léna: Jean-Marie Mariotti (1955–1998)

93, 47Personnel Movements 93, 47Atlas of the Northern Sky 94, 31ESO Director General to Become President of

AUI 94, 39First Announcement of the ESO Workshop on

“Black Holes in Binaries and Galactic Nuclei– Their Diagnostics, Demography, andFormation” 94, 39

VLT Opening Symposium 94, 40Personnel Movements (1 October – 31 De-

cember 1998) 94, 41List of Scientific Preprints (September – De-

cember 1998) 94, 41ESO Workshop Proceedings Still Avail-

able 94, 41

43

A U T H O R I N D E XA

M. Albrecht: The VLT Data Volume 93, 21P. Andreani: Using SEST to Probe the Geom-

etry of the Universe 91, 35 I. Appenzeller, K. Fricke, W. Fürtig, W. Gässler,

R. Häfner, R. Harke, H.-J. Hess, W. Hum-mel, P. Jürgens, R.-P. Kudritzki, K.-H. Man-tel, W. Meisl, B. Muschielok, H. Nicklas, G.Rupprecht, W. Seifert, O. Stahl, T. Szeifert,K. Tarantik: Successful Commissioning ofFORS1 – the First Optical Instrument onthe VLT 94, 1

M. Arnaboldi, M. Capaccioli, D. Mancini, P.Rafanelli, R. Scaramella, G. Sedmak, G.P.Vettolani: VST: VLT Survey Telescope 93,30

T. Augusteijn: 2.2-m Telescope Upgrade: aStatus Report 94, 12

BD. Baade et al.: The Wide Field Imager for the

2.2-m MPG/ESO Telescope: a Preview93, 13

P. Bonifacio and P. Molaro: Upgraded NTTProvides Insights Into Cosmic BigBang 92, 26

L. Binette, B. Joguet, J.C.L. Wang, G. MagrisC.: NTTArchives: the Lyman α Profile of theRadio Galaxy 1243+036 Revisited 92,37

H. Böhringer, L. Guzzo, C.A. Collins, D.M. Neu-mann, S. Schindler, P. Schuecker, R. Crud-dace, S. Degrandi, G. Chincarini, A.C.Edge, H.T. Macgillivray, P. Shaver, G. Vet-tolani, W. Voges: Probing the CosmicLarge-Scale Structure with the REFLEXCluster Survey: Profile of an ESO KeyProgramme 94, 21

J. Brewer and J. Andersen: Improving ImageQuality at the Danish 1.54-m Tele-scope 92, 18

I. Burud, F. Courbin, C. Lidman, G. Meylan, P.Magain, A.O. Jaunsen, J. Hjorth, R.Ostensen, M.l. Andersen, J.W. Clasen, R.Stabell, S. Refsdal: RXJ0911.4+0551: AComplex Quadruply Imaged GravitationallyLensed QSO 92, 29

CR. Chini and E. Krügel: Cold Dust in Galaxies

91, 37F. Combes and T. Wiklind: Molecular Lines in

Absorption at High Redshift 91, 29F. Comerón, R. Neuhäuser, A.A. Kaas: The

First X-ray Emitting Brown Dwarf 94, 28J.G. Cuby and R. Gilmozzi: VIMOS and NIR-

MOS: Status Report 91, 16

DL. Da Costa, E. Bertin, E. Deul, T. Erben, W.

Freudling, M.D. Guarnieri, I. Hook, R. Hook,R. Mendez, M. Nonino, L. Olsen, l. Prandoni,A. Renzini, S. Savaglio, M. Scodeggio, D.Silva, R. Slijkhuis, A. Wicenec, R. Wich-

mann, C. Benoist: The ESO Imaging Sur-vey: Status Report and Preliminary Results91, 49

L. da Costa and A. Renzini: ESO ImagingSurvey: Update on ElS-deep and the HubbleDeep Field South 92, 40

S. D’Odorico: The New Direct CCD ImagingCamera at the NTT. SUSI2 gets the firstFIERA CCD controller and a mosaic of two2k × 4k, 15µ pixel, thinned, anti-reflectioncoated EEV chips 91, 14

M. Dolenski, A. Micol, B. Pirenne, M. Rosa:How the Analysis of HST EngineeringTelemetry Supports the WFPC2 AssociationProject and Enhances FOS CalibrationAccuracy 93, 23

M. Dolenski, A. Micol, B. Pirenne: HST ArchiveServices Implemented in Java 93, 26

GR. Giacconi: The Role of ESO in European

Astronomy 91, 1R. Giacconi: First Light of the VLT Unit

Telescope 1 92, 2Ph. Gitton and L. Noethe: Tuning of the

NTTAlignment 92, 15R. Gredel and P. Leisy: EFOSC2 and VLT

Autoguider Commissioned at the 3.6-mTelescope 91, 24

S. Guisard: Image Quality of the 3.6-mTelescope (Part VlII). New Seeing Recordof 0.47″ 91, 24

HO.R. Hainaut: News from the NTT 93, 16O.R. Hainaut: News from the NTT 94, 11C. Henkel, Y.-N. Chin, R. Wielebinski, R.

Mauersberger: Cool Gas in SouthernGalaxies 91, 45

KB. Koehler: ESO and AMOS Signed Contract

for the VLTI Auxiliary Telescopes 93, 11 B. Koehler, F. Koch: UT1 Passes “With

Honour” the First Severe Stability tests forVLTI 93, 11

M. Kürster: CES Very Long Camera Installed92, 18

M. Kürster: Performance of CES 3.6-m FibreLink and Image Slicers 94, 12

LP.O. Lagage, Y. Rio, J.W. Pel, H. Toisma:

VISIR at PDR 91, 17B. Leibundgut, G. de Marchi, A. Renzini:

Science Verification of the VLT UnitTelescope 1 92, 5

Pierre Léna: Jean-Marie Mariotti (1955–1998)93, 47

J. Lequeux: Carbon Monoxyde in the Magel-lanic Clouds 91, 41

C. Lidman: SOFI Receives its First Users 93,16

MC. Madsen, R. West: Sea & Space – A

Successful Educational Project for Europe’sSecondary Schools 93, 44

C. Madsen: Award for the ESO VLTProject 94, 38

G. Mathys: News from the NTT 91, 21G. Mathys: News from the NTT 92, 14A. Micol, D. Durand, S. Gaudet, B. Pirenne:

HST Archive News: On the Fly Recalibration(OTF) of NICMOS and STIS Data 93, 27

A. Micol, B. Pirenne: HST Archive News:WFPC2 Associations 93, 28

R.P. Mignani, S. Mereghetti, C. Gouiffes andP.A. Caraveo: Timing, Spectroscopy andMulticolour Imaging of the Candidate OpticalCounterpart of PSR B1509–58 94, 25

A. Moorwood, J.-G. Cuby and C. Lidman: SOFISees First Light at the NTT 91, 9

A. Moorwood, J.-G. Cuby, P. Biereichel, J.Brynnel, B. Delabre, N. Devillard, A. VanDijsseldonk, G. Finger, H. Gemperlein, R.Gilmozzi, T. Herlin, G. Huster, J. Knudstrup,C. Lidman, J.-L. Lizon, H. Mehrgan, M.Meyer, G. Nicolini, M. Petr, J. Spyromilio, J.Stegmeier: ISAAC Sees First Light at theVLT 94, 7

NR. Neuhäuser, H.-C. Thomas, F.M. Walter:

Ground-Based Detection of the IsolatedNeutron Star RXJ185635-3754 at V = 25.7Mag with the Upgraded NTT 92, 27

B. Nordström: A Fresh Look at the Future: “LaSilla 2000++” 93, 42

L.-Å. Nyman and A. Tieftrunk: SEST Upgrades91, 28

OA. Otárola, G. Delgado, R. Booth, V. Belitsky,

D. Urbain, S. Radford, D. Hofstadt, L.Nyman, P. Shaver, R. Hills: European SiteTesting at Chajnantor 94, 13

PR.G. Petrov, F. Malbet, A. Richichi, K.-H. Hof-

mann: AMBER, the Near-lnfrared/Red VLTIFocal Instrument 92, 11

B. Pirenne, M. Albrecht, B. Leibundgut: ESOand ST-ECF Archive News 93, 20

B. Pirenne, M. Albrecht: Using DVD Technologyfor Archiving Astronomical Data 93, 22

B. Pirenne, B. McLean, B. Lasker: ST-ECFParticipation in the GSC-II GenerationProject 93, 25

QP. Quinn, J. Breysacher, D. Silva: First Call for

Proposals 92, 9

RA. Renzini: Après EIS 91, 54M. Rubio, G. Garay, R. Probst: Molecular Gas

in 30 Doradus 93, 38

SJ. Sanner, M. Geffert, H. Böhnhardt, A. Fiedler:

Astrometry of Comet 46P/Wirtanen at ESO:Preparation of ESA’s ROSETTA Mission92, 33

M. Sarazin: Astroclimate During ScienceVerification 93, 2

P.A. Shaver and R.S. Booth: The LargeSouthern Array 91, 26

D. Silva: The NTT Service Observing Pro-gramme: Period 60. Summary and LessonsLearned 92, 20

S. Stanghellini and A. Michel: Performance ofthe First Two Beryllium Secondary Mirrorsof the VLT 94, 10

M. Sterzik: 3.6-m Telescope Passes MajorUpgrade Milestones 93, 17

TM. Tarenghi, P. G ray, J. Spyromilio, R. Gil-

mozzi: The First Steps of UT1 93, 4A. R. Tieftrunk, S. T. Megeath: Star Forma-

tion Toward the “Quiescent” Core NGC6334 I(N) 93, 36

UM.-H. Ulrich: Observations with Adaptive

Optics: Getting There 94, 32

VM.-P. Véron, G. Meylan: 6th ESO/OHP Sum-

mer School in Astrophysical Observations93, 43

WR.M. West: VLT First Light and the Public 92, 4

44

ESO, the European Southern Observatory,was created in 1962 to “… establish andoperate an astronomical observatory in thesouthern hemisphere, equipped with pow-erful instruments, with the aim of further-ing and organising collaboration in astron-omy …” It is supported by eight countries:Belgium, Denmark, France, Germany, Italy,the Netherlands, Sweden and Switzerland.ESO operates at two sites. It operates theLa Silla observatory in the Atacama desert,600 km north of Santiago de Chile, at 2,400m altitude, where fourteen optical tele-scopes with diameters up to 3.6 m and a15-m submillimetre radio telescope (SEST)are now in operation. In addition, ESO isin the process of building the Very LargeTelescope (VLT) on Paranal, a 2,600 mhigh mountain approximately 130 km southof Antofagasta, in the driest part of theAtacama desert. The VLT consists of four8.2-metre and several 1.8-metre tele-scopes. These telescopes can also be usedin combination as a giant interferometer(VLTI). “First Light” of the first 8.2-metretelescope (UT1) occurred in May 1998. UT1will be available on a regular basis for as-tronomical observations from April 1999 on.Over 1000 proposals are made each yearfor the use of the ESO telescopes. The ESOHeadquarters are located in Garching,near Munich, Germany. This is the scien-tific, technical and administrative centre ofESO where technical development pro-grammes are carried out to provide the LaSilla and Paranal observatories with themost advanced instruments. There arealso extensive astronomical data facilities.In Europe ESO employs about 200 inter-national staff members, Fellows and As-sociates; in Chile about 70 and, in addition,about 130 local staff members.

The ESO MESSENGER is published fourtimes a year: normally in March, June,September and December. ESO also pub-lishes Conference Proceedings, Preprints,Technical Notes and other material con-nected to its activities. Press Releases in-form the media about particular events. Forfurther information, contact the ESO Edu-cation and Public Relations Department atthe following address:

EUROPEAN SOUTHERN OBSERVATORY Karl-Schwarzschild-Str. 2 D-85748 Garching bei München Germany Tel. (089) 320 06-0 Telefax (089) 3202362ips@eso.org (internet) URL: http://www.eso.org

The ESO Messenger:Editor: Marie-Hélène DemoulinTechnical editor: Kurt Kjär

Printed by Druckbetriebe Lettner KGGeorgenstr. 84D-80799 MünchenGermany

ISSN 0722-6691

ContentsI. Appenzeller, K. Fricke, W. Fürtig, W. Gässler, R. Häfner, R. Harke,

H.-J. Hess, W. Hummel, P. Jürgens, R.-P. Kudritzki, K.-H. Mantel,W. Meisl, B. Muschielok, H. Nicklas, G. Rupprecht, W. Seifert, O. Stahl,T. Szeifert, K. Tarantik: Successful Commissioning of FORS1 – theFirst Optical Instrument on the VLT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

A. Moorwood, J.-G. Cuby, P. Biereichel, J. Brynnel, B. Delabre, N. Devillard,A. Van Dijsseldonk, G. Finger, H. Gemperlein, R. Gilmozzi, T. Herlin,G. Huster, J. Knudstrup, C. Lidman, J.-L. Lizon, H. Mehrgan, M. Meyer,G. Nicolini, M. Petr, J. Spyromilio, J. Stegmeier: ISAAC Sees First Lightat the VLT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

News from VLT Science Verification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

TELESCOPES AND INSTRUMENTATION

S. Stanghellini and A. Michel: Performance of the First Two BerylliumSecondary Mirrors of the VLT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

THE LA SILLA NEWS PAGE

O.R. Hainaut: News from the NTT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11T. Augusteijn: 2.2-m Telescope Upgrade: a Status Report . . . . . . . . . . . . . . . 12M. Kürster: Performance of CES 3.6-m Fibre Link and Image Slicers . . . . . . . 12

NEWS FROM THE LSA

A. Otárola, G. Delgado, R. Booth, V. Belitsky, D. Urbain, S. Radford, D. Hof-stadt, L. Nyman, P. Shaver, R. Hills: European Site Testing at Chajnantor . 13

REPORTS FROM OBSERVERS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

H. Böhringer, L. Guzzo, C.A. Collins, D.M. Neumann, S. Schindler,P. Schuecker, R. Cruddace, S. Degrandi, G. Chincarini, A.C. Edge,H.T. Macgillivray, P. Shaver, G. Vettolani, W. Voges: Probing the CosmicLarge-Scale Structure with the REFLEX Cluster Survey: Profile of an ESOKey Programme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

R.P. Mignani, S. Mereghetti, C. Gouiffes And P. A. Caraveo: Timing, Spec-troscopy and Multicolour Imaging of the Candidate Optical Counterpart ofPSR B1509–58 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

F. Comerón, R. Neuhäuser, A.A. Kaas: The First X-ray Emitting Brown Dwarf 28Call for Ideas for Future Public Imaging Surveys . . . . . . . . . . . . . . . . . . . . . . . 31Atlas of the Northern Sky . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

OTHER ASTRONOMICAL NEWS

M.-H. Ulrich: Observations with Adaptive Optics: Getting There . . . . . . . . . . . 32C. Madsen: Award for the ESO VLT Project . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

ANNOUNCEMENTS

ESO Director General to Become President of AUI . . . . . . . . . . . . . . . . . . . . . 39First Announcement of the ESO Workshop on “Black Holes in Binaries and

Galactic Nuclei – Their Diagnostics, Demography, and Formation” . . . . . . . 39VLT Opening Symposium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40Personnel Movements (1 October – 31 December 1998) . . . . . . . . . . . . . . . . 41List of Scientific Preprints (September–December 1998) . . . . . . . . . . . . . . . . . 41ESO Workshop Proceedings Still Available . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

MESSENGER INDEX 1998 (Nos. 91–94) . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42