No. 60 - -.June 1990 --, Planning the VLTVLT Interferometer J. M. BECKERS, ESO 1. The VlVLTT Interferometer: VLTVLT Reports 44 and 49), the interfero­interfero- approved YLTVLT implementation. P. LemaLena One of the Operating Modes metric mode of the VLVLTT was indudedincluded in described the concept and planning torfor ofoftheVLT the VLT the VLTVLT proposal, and accepted inin theIhe the inlerferometricinterferometric mode of the VLTVLT at 1.1.11 Itsfts Context Adaptive Optics at the ESO 3.6-m Telescope The Very Large Telescope has three differenldifferent modes of being used. As four separate 8-metre telescopes it provides theIhe capabililycapability of carrying out in parallel tourfour different observing programmes, each with a sensitivitysensilivity which matches that of the other most powerful ground­ground- based telescopes available. In the sec­sec- ond mode the light of the four tele-lele­ scopes is combined in a single imageimage making ilit inin sensitivity the most powerful telescope on earth, almost 16t6 metres in diameter if the light losses in the beam eombinationcombination canean be kept low. In the third mode Ihethe light of the four tele-tele­ scopes is combined coherenlly,coherently, allow­allow- This faise-colourfalse-colour photo illustrates the dramatiedramatic improvement in image sharpness whiehwhich is ing interferomelrlcinterferometric observations with the obtained with adaptive optiesoptics at the ESO 3.6·m3.6-m telescope. seeSee also the artielearticle on page 9. unparalleled sensitivity resulting fromtrom I1It shows the 5.5 magnitudemagnilude star HR 6658 in the galacticga/aeUe eluslercluster Messier 7 (NGC 6475), as the 8-metre apertures. In this mode theIhe observed in the infrared L·bandL-band (wavelength 3.5pm),3.Sllm), without ("ullcorrecled",("uncorrected", left) alldand with angular resolulionresolution is determined by the "corrected","corrected, right) the "VL"VLTT adaptive optics prototype" switched on. The diameter of the distance between the telescopes (up to10 uneorrecteduncorrected image is about 0.8 arcseconds, eorrespondingcorresponding to the instantaneousinstanlaneous "seeing"seeing"9 diskodisk. 120 metres), rather than by the resolu­resolu- When corrected, the image sharpness increasesinereases nearly fourfold; the diameter is now only 0.22 tion determined by the individual tele-tele­ arcsec. This corresponds to lhethe diffraction Umitlimit at this wavelength; the diffraction rings are weilwell visible. The improved sharpness reveals that the star is double; the angular distance between scopes (set by the atmosphericalmospheric seeing, visible. The improved sharpness revea/s that t!Je star Is double; the angular distanee betw8en the two eomponentscomponents is 0.38 areseconds.arcseconds. or by the diffraction limit of single Although it is not evident on this picturepieture in whiehwhich theIhe intensity sealesscales have been normalized to 8-metre apertures while using adaptive the same level, theIM eentralcentral intensity of thet!Je eorrecledcorrected image is much higher than that of Ihethe optics or speckle interferometryj.interferometry). uncorrected. SyBy coneentratingconcentrating Ihethe light belter,better, Ihethe effideneyefficiency of the leleseopetelescope is eorrespond­correspond- Following an in-depth study by the VLTVLT ingly inereased.increased. This means that shorter exposure times are possible or thaithat fainter objects can InterferometryInlerferometry Working Group under the be observed than before. chairmanship of P. LenaLfma (published in thatthat timetime inin anan earlierearlier articlearticle inin thethe thethe change-overchange-over fromfrom thethe useuse ofof thethe tiontion planplan andand proposalproposal viewview thethe de-de­ MessengerMessenger(P. (P. Lena,Lena, TheThe MessengerMessenger53, 53, sub-arraysub-array toto thethe fullfull VLTlVLTI requiresrequires aa velopmentvelopment ofof interferometryinterferometry withwith thethe 53,53, 1987).1987). minimumminimum ofof steps.steps. VLTlVLTI toto bebe aa gradual,gradual, step-by-stepstep-by-step one.one. SinceSince thethe acceptanceacceptance ofof thethe VLTVLT AlthoughAlthough suchsuch anan integratedintegrated ap-ap­ TheThe plan,plan, however,however, differsdiffers fromfrom thethe origi-origi­ proposal,proposal, thethe VLTVLT finalfinal definitiondefinition andand proachproach maymay appearappear obviousobvious toto thethe read-read­ nalnal proposalproposal inin somesome aspects:aspects: (i)(i) itit design,design, includingincluding thethe finalfinal sitesite testingtesting er,er, itit isis byby nono meansmeans anan uncontroversial,uncontroversial, arguesargues stronglystrongly forfor aa configurationconfiguration ofof thethe whichwhich willwill soonsoon leadlead toto thethe VLTVLT sitesite obviousobvious philosophyphilosophy toto everyoneeveryone inin thethe 8-metre8-metre telescopestelescopes whichwhich isis moremore op-op­ choice,choice, has been rapidly proceeding. interferometryinterferometry community.community. There firstfirst isis timizedtimized forfor interferometryinterferometry (see(see sectionsection TheThe implementation planning of thethe in-in­ thethe skepticismskepticism by many about thethe actual 3.1), (ii)(ii) itit provides forfor thethe expansibility terferometricterferometric mode isis an essential part availability of thethe 8-metre telescopestelescopes forfor of thethe array withwith more auxiliary tele-tele­ ofof that.that. lnterferometrylnterferometry places itsits ownown interferornetricinterferometric imaging.imaging. Certainly, thesethese scopesscopes andand delaydelay lineslines (to(to be provided requirementsrequirements onon thethe design and location largelarge telescopestelescopes willwill not be available forfor by additionaladditional contributionscontributions by ESO of thethe 8-metre telescopes,telescopes, onon thethe sitesite a significant amount of thethe timetime forfor inter-inter­ member states),states), and (iii)(iii) itit provides forfor choice and development, and in thethe de-de­ ferometricferometric imagingimaging until thethe astronomy thethe incorporationincorporation of adaptive optics inin finitionfinition of thethe observatory infrastructure. community wants toto use them for that.that. thethe auxiliary telescopes.telescopes. InIn addition toto thethe 8-metre telescopes, Before their use forfor interferometricinterferometric im-im­ thethe VLTVLT includesincludes twotwo smaller movable aging will successfully compete inin timetime 1.3 Why Large Aperture 2-metre-class telescopes (the so-called with other uses, a "user-friendly" capa-capa­ Interferometry?Interferometry? auxiliary telescopes), whose design and bility for interferometric imaging with thethe configuration has toto proceed in parallel VLTlVLTI has toto be demonstrated. That will In thisthis context thethe question is some-some­ with thethe others. To aid in thisthis planning a have toto be done with thethe auxiliary tele-tele­ timestimes asked on what is really gained by VLTVLT lnterferometrylnterferometry Panel was formed, scope sub-array, which will also exploit going toto largelarge aperture interferometers.interferometers. whose members represented thethe exper-exper­ ways of reaching thethe maximum possible Doesn't an array of many small tele-tele­ tisetise inin interferometryinterferometry inin thethe ESO sensitivity. Itlt is our conviction that thisthis scopes do as well,weil, or even better? The member countries'.countries1. This report can be done, and thatthat thethe resulting user answer to this question is important be-be­ summarizes the implementation plan of demand for the use of the 8-metre tele-tele­ cause it determines the part of as-as­ the panel for the so-called VLTVL T Inter-Inter­ scopes for interferornetricinterferometric imaging will tronomy at which the VLTlVLTI will be par-par­ ferometer (or VLTI)VLTI) as proposed to ESO. follow. Second, the design of the ticularly useful and for which it should This plan is out of necessity incomplete 8-metre telescopes cannot be op-op­ be optimized. Although an increase in since the site for the VLTVLTremains remains to be timized for interferometry alone. There-There­ collecting area in astronomy is generally chosen. fore, compromises are necessary which thought of as giving an increase in sen-sen­ ripple through to the design of the entire sitivity, this is by no means obvious for VLTI, including the sub-array of auxiliary interferometricinterferornetric imaging. It has been 1.2 Philosophy Followed in the 1.2 Philosophy Followed in the telescopes because of the philosophy argued in fact that little is gained in Implementation Plan argued in fact that little is gained in Implementation Plan described above. Certainly, an inter­inter- limiting sensitivity by the increased Right from the beginning the panel ferometer designed without the con-con­ aperture in the case of multi-speckle, took as the basis for its studies the straints associated with the VLVLTT (site, broadband observations. This is the desire to make use of the unique oppor­oppor- fixed telescopes requiring the use of case at short wavelengths (up to the tunity provided by the presence on one delay lines, alt-az mounts, limited Johnson M band without the use of site of four identical, state-of-the-art number of telescopes, etc.) would in adaptive optics, up to the H band with 8-metre telescopes to do interferometric some aspects be more powerful than the use of adaptive optics),optics). The combi­combi- imaging. Its definition of the VLVLTlTI is thus