messenger no154

8/18/2019 Messenger No154

1/72

The Messenger

No. 154 – December 2013

P a r a n

a l i n s t r u m e n t a t i o n p r o

g r a m m e

F o c u s o n E S O

P u b l i c S u r v e

y s

R e s o

l v i n g A G N

w i t h M I D I


2/72


3/72

3The Messenger 154 – December 2013

instruments. The second phase is dic-tated by the strategy of how the VLT willbe used in the E-ELT era.

Phase 1 (Projects initiated before 2018/ deployed before ~ 2025). There is no indication that the size of theParanal user community will decrease.On the contrary, new Member States may join ESO, increasing the pressure onthe Paranal facilities. Consequently thescientic use and output of Paranalinstruments should be optimised. It isimportant to preserve a balance betweenspecialised instruments and workhorseinstruments, with the latter covering awide range of scientic interests.

Phase 2 (Long-term opportunities in theE-ELT era, af ter ~ 2025). This phase is still re latively open anddifferent scenarios can be envisaged. The E-ELT will be fully operational andastronomical research with 8-metre-class telescopes may evolve towards amodel where a large fraction of the timeis devoted to dedicated experimentsand large collaborative projects. In thiscontext the four VLT Unit Telescopestogether could provide a unique oppor-tunity to dedicate up to ~ 1200 nightsper year to a single problem. Thisapproach could open up new perspec-tives in astronomical research. The lasttwo instruments of the decade (deployedin 2018/2019), should be fully integratedinto this long-term perspective. Theirselection will occur after a careful reec-tion on the scientic use and role of the VLT in the E-ELT era. To this purpose,several scientic conferences will be heldin the coming years to direct the choicesand nalise the strategy and its imple-mentation.

Programmatic drivers

The instrumentation development planfollows from consideration of a numberof basic drivers:

Paranal and E-ELT The E-ELT will be an additional telescopeat the Paranal Observatory, and the

strengths of each unit should be maxim-ised. Synergy and the ability to comple-ment E-ELT capabilities are thereforeimportant criteria for the VLT.

Paranal, HST and JWSTBy 2018 the Hubble Space Telescope(HST) will most likely no longer be inoperation, and the James Webb Space Telescope (JWST ) will be about toenter operations. HST capabilities thatwill be unavailable include ultraviolet(UV) spectroscopy and high-resolutionimaging in the B- to R-bands. An instru-ment able to provide diffraction-limitedobservations in the B- to R-bands overa sufciently large eld could recoveran important part of the missing parame-ter space. Complementarity betweenthe VLT and the JWST in the areas ofhigh resolution spectroscopy, observationof bright sources, diffraction-limitedobservation at short wavelengths, exibleoperations, wide wavelength coverageand use of wide eld can be mentioned.It may also be advantageous to providesome overlapping capabilities with JWST.

Paranal and ground-based observatories The relationship of Paranal with otherground-based observatories (including ALMA) has sti ll to be discussed in depth.In general, the Paranal choices will bedriven by the scientic requests of theESO community rather than by the devel-opments of its competitors.

Maximisation of efciency/optimal useof observing timeOptimisation can be achieved by concen-trating on three main aspects: improvedefciency (throughput and duty cycle);extending the spectral coverage; explor-ing the possibilities for sharing the availa-ble foci. This goal could include the con-cept (new for the VLT) of instrumentsdesigned to be exchanged with a regularcadence.

Instrument development duration The typical development time for secondgeneration VLT instruments has beenalmost ten years from the time of con-ception. This long lead time should notbe assumed to be inevitable, and theprogramme could develop instrumentson shorter construction times if thisbecomes an agreed goal. One interestingpossibility would be to create a newclass of visitor instrument, operated bythe construction team, but also includ-ing proposals from the community atlarge (in the manner of the VLTI instru-ment PIONIER).

Focus occupancyWith the arrival of ESPRESSO in 2016,all VLT/I foci will be occupied, includingthe incoherent combined focus. Someinstruments (e.g., ISAAC and MIDI) willhave been decommissioned as earlyas 2013–2014 and replaced by secondgeneration instruments (SPHERE, MUSEand GRAVITY). Each time a new instru-ment is accepted, the instrument to bedecommissioned will be identied on thebasis of a grid of criteria that includes:scientic potential, complementarity withnew instruments, instrument status andfuture perspectives.

Role of La SillaIt is clear that today the success of4-metre-class telescopes is often linkedto the ability to occupy scientic niches.HARPS at the ESO 3.6-metre telescopeis a good example of such a successstory. The specic added value of theLa Silla 4-metre-class telescopes for ESOcan be summarised:– La Silla continues to be a competitive

site in the southern hemisphere pro-viding unique opportunities to its users;

– the ESO community continues torequest the ESO 3.6-metre telescopeand the New Technology Telescope(NTT) at reasonable to high oversub-scription rates and both telescopescontinue to produce good publicationrates (105 refereed publications fromthe NTT in 2012, 69 from the ESO 3.6-metre);

– the ESO 3.6-metre telescope and theNTT are maintained to VLT technicalstandards and provide excellent imagequality and efciency at negligible tech-nical down time;

– a minority of Member States haveaccess to national 4-metre-class tele-scopes;

– La Silla provides the opportunity todedicate a 4-metre-class telescope toone, or a few, scientic questions;

– 4-metre-class telescopes with state-of-the-art (workhorse) instrumentationrelease pressure on the observing timeat the VLT (and in the future, possiblyfor the E-ELT).

Considering that the current NTT instru-mentation is reaching the end of its lifecycle (EFOSC2 went into operation in1990, SOFI in 1998), ESO will launch acall in 2014 for a new instrument for the


4/72

4 The Messenger 154 – December 2013

wavelength range by a large factor. Anupgrade that considers the installationof a set of cross-dispersers and newdetectors has passed Phase A and, afterpositive STC recommendation, is nowin the design and construction phase(Oliva et al., 2012). It will answer a numberof scientically pressing questions, andwill, in addition, satisfy several of theabove considerations (such as comple-mentarity with JWST and improvement ofefciency).

MOONS and 4MOST

The proposal to bui ld a new, powerfulmulti-object spectrograph (MOS) hasbeen strongly endorsed by the ESO com-munity and advocated in several instancesby the STC. After a call for ideas, twocompetitive MOS Phase A studies wereawarded: 4MOST (de Jong et al., 2011)and MOONS (Cirasuolo et al., 2011).

MOONS is a near infrared facility (0.8–1.8 μm) which can host up to 1000 bresat the Nasmyth focus of the VLT. Theeld of view is about 500 square arc-minutes. It can operate either at lowerresolution (R ~ 5000) or at higher resolu-tion (R ~ 20 000) in two selected spectralregions.

4MOST is proposed for the 4-metre VISTA telescope, with a eld of view ofmore than 3 square degrees. It will hostup to 2400 bres and will work in theoptical (0.3–0.9 μm). Sixteen hundredbres will feed two lower resolution spec-trographs (R ~ 5000), with 800 bresto two higher resolution spectrographs(R ~ 20 000).

These two instruments are largelycomplementary in almost all aspects:spectral coverage, telescope used, eldof view and scientic aims. Given theoutstanding science cases presented bythe two consortia, the enormous rangeof applications of large eld spectroscopyand the strong push by the communityto increase ESO’s MOS capabilities,together with the strong complementaritywith JWST and E-ELT, both instrumentshave been recommended for designand construction by STC. The work for

MOONS has star ted in 2013 and 4MOSTwill start in 2014.

NTT to be built in the community. Thisnew instrument could replace eitherSOFI or EFOSC2 or both, and would beavailable to the ESO community for50% of the time until 2021. Additionalobserving time with the new instrumentwill be available for interested groupsthrough the co-funding of NTT opera-tions.

The NTT call wil l be open for both spe-cialised instruments, taking advantageof the large amount of dedicated ob-serving time, as well as state-of-the-artworkhorse instruments addressing broadneeds within the ESO community. Suchan instrument is required to be at negli-gible cost to ESO.

Instrument denition and procurementprocedure

Scientic input for the new instruments isprovided to the instrumentation pro-gramme manager through:– the “Paranal in the E-ELT era” white

paper, as well as other inputs from the VLT programme scientist;

– the community, by either contacting theinstrumentation programme managerdirectly, via the STC, or via ad hoc sci-entic conference(s).

Each proposal will be scientically evalu-ated by the VLT programme scientist.In order to ensure community input to thedenition of the Paranal instruments, sci-entic workshops will be organised toaddress the scientic needs for the VLT inthe next decade. The emphasis will beon 8-metre telescope science rather thaninstruments or technological concepts. These workshops, organised in theperiod 2013–2017, will dene the instru-mental capabilities to be developed after~ 2018.

A working group of about 15 people(ve from ESO, ve composed of STCmembers and ve community experts)will evaluate the best sequence in whichto deploy the 2015–2018 projects. Anon-exhaustive list of instrument options,which has emerged so far from the dif-ferent inputs, is presented in the followingsections.

After the var ious inputs have been col-lected and elaborated, a proposalconsisting of the top-level characteristicsfor the instruments will be presentedto the STC. Once the concept has beenrecommended a call for tenders forPhase A study will follow.

The Paranal instrumentation programmewill not be static, and must be able toreact to the evolving scientic and tech-nological landscape and to re-assignpriorities. New proposals will be evalu-ated by the programme manager, in col-laboration with ESO management andthe STC, against the existing plan. Acceptance of a new project may resulteither in cancelling/de-scoping or re-phasing planned projects. A similar evalu-ation will be made if one of the runningprojects requests a substantial increasein the allocated resources.

New instruments for the VLT

Following a series of Phase A studiesand recommendations by the STC, thefollowing new instruments are now inprocess.

CUBESIn UV spectroscopy from the ground(i.e., 300–380 nm spectral range), alarge increase of efciency with respectto the existing instruments (UVES and X-shooter) is possible. In addit ion, thisspectral range complements that of theE-ELT and JWST. An efcient UV spec-trograph can cover a broad sciencecase and could be a world-leading instru-ment for many years to come. Locatedat the Cassegrain focus, it could be easilyexchanged. The CUBES concept will bedeveloped by a consortium of Brazilianinstitutes and ESO. The project haspassed Phase A review and has beenrecommended by the STC. The detaileddesign is ongoing, and constructionwill commence following the raticationof Brazilian accession to ESO.

CRIRES upgradeCRIRES is equipped with a pre-disperserand currently delivers a fraction of oneechelle order per observation. A cross-disperser could increase the simultaneous

Telescopes and Instrumentation Pasquini L. et a l., Paranal Instrumentation Programme


5/72


Potential new instruments for the VLT/I

After examining the current complementof Paranal instruments at the telescope,or in construction, a number of potentialdevelopments can be identied, whichare listed below. This list is not intendedto be exhaustive.

Workhorse instrument to complement/ support FORS2 and X-shooterFORS2, X-shooter and ISAAC (and alsoEFOSC at the NTT) are among themost popular and productive ESO instru-ments. They are typical workhorses andthe user pressure on them is very high.It is important that ESO preserves thisclass of instrument. With the decommis-sioning of ISAAC, infrared spectroscopyin the 2.4–5 μm regime will no longerbe available. Should the new workhorsebe a multi-function mul ti-wavelengthinstrument? Or a copy (perhaps slightlymodied) of one of the existing, mostrequested instruments? Such questionswill be debated by the ESO/STC/commu-nity working group.

New Instrument for the AOFIn answer to the STC request for a planfor AO instruments at the VLT, ESOhas proposed a development in twophases: ERIS, that will follow-up NACOand feed SPIFFI, the SINFONI spectro-graph; a new, ambitious instrument,still to be decided, to fully exploit thepotential of the AOF, in the focus occu-pied by GRAAL and HAWK-I. A highStrehl B- to R-band imager would be oneattractive possibility. A multi-IFU, AO-assisted, large eld spectrograph wouldalso be unique, and its scientic meritsshould be studied.

In either case, the instrument may requirea considerable amount of research anddevelopment. The scientic discussionabout a new AO-assisted instrument ofthis type should start soon.

New VLTI instrument

The VLTI wil l continue to provide thehighest angular resolution, even in the

E-ELT era. The rising demand for imagingcapability of stellar sur faces, close cir-

cumstellar environments and extra-galactic sources sets a clear path forthe VLTI medium-term developmentplan. PIONIER, GRAVITY, MATISSE andthe second generation fringe trackerare, and will be, the immediate answersto that request. The continuous andsuccessful effor t to improve the VLTI’srobustness and performance will beessential too.

However, improving the spectral cover-age (visible to mid-infrared) and the imag-ing capability of VLTI should remain ahigh priority in the years to come. PIONIERalready provides this and GRAVITY willprovide observing modes close to themost demanded AMBER ones, but withgreater sensitivity and much improvedFourier uv-plane coverage.

While it seems premature to star t a newproject given the enormous ongoingeffort to complete and operate PRIMA,GRAVITY and MATISSE, some avenuesto be explored for the VLTI in the comingyears include:1) Securing the continuity of PIONIER and

offering it to the community;2) Continuing to offer a visitor focus at the

VLTI;3) Exploring the six-telescope imaging

capabilities of VLTI with the existinginfrastructure.

Potential VLT instrument upgrades

Even if most of the VLT/I instruments willbe new or recently upgraded, the 15 yearsof VLT experience demonstrate that thereare frequent requests for upgrades(mostly of detectors) and that these haveserved the community very well.Upgrades under consideration are:– X-shooter: Two proposals to upgrade

X-shooter have been submitted andhave been evaluated.

– FORS2: A proposal to upgrade theFORS2 detector is being prepared. The use of a 4kx4k pixel CCD detectorwould bring substantial operationalbenets.

– SPHERE: The deformable mirror is for-mally below specication, and areplacement could be needed if its per-formance deteriorates.

All major upgrades will be treated as anyother project, and compared to runningor planned instruments in order to decidepriorities. It must be clear that startingone project per year implies that either anew instrument or a major upgrade canbe initiated, but not both.

Potential new instruments for La Silla

3.6-metreIn 2014 HARPS will be equipped with theLaser Frequency Comb (LFC), but willbe out-performed by ESPRESSO at the VLT after ~ 2017. However HARPShas the advantage of using a dedicatedtelescope and of having built up a longtime-series of observations; it shouldbe used for the sources that do not needan 8-metre collecting area. HARPS ishighly requested, and its future demandwill also depend on the fate of space mis-sions. It is worth recalling that exoplanetscience is a young and expanding eld.

NTT The new instrument for the NTT (seeabove) could either be a dedicated instru-ment or a multi-function workhorse.

An exciting approach could be to com-plement HARPS at the ESO 3.6-metrewith a near-infrared planet-nder atthe NTT, matching the RV precision ofHARPS. Several observatories areplanning instruments of this kind in thenorthern hemisphere. High velocity pre-cision could make it unique.

An X-shooter-type instrument for theNTT could be an interesting alternativeto EFOSC2 plus SOFI.

Roadmap

Starting from 2013, the resources dedi-cated to E-ELT instrumentation willprogressively increase. This will imply aprogressive decrease in the resourcesavailable for the Paranal instrumentationprogramme to a new level that cansustain the “one new start per year” plan. Table 1 shows the proposed t imetable.For all instruments, one year of Phase Ais foreseen and a development time


6/72


of ve years. This is on the short side,but not unrealistic. Figure 2 shows theParanal instrumentation and the projectdevelopment in 2019 according to thepresent plan. In a resource-constrainedenvironment, the beginning of new projects will also have to be subject to satis-factory completion of existing projects.If existing projects run late, the new oneswill be re-planned accordingly.

References

Cirasuolo, M. et al. 2011, The Messenger, 145, 11de Jong, R. et al. 2011, The Messenger, 145, 14Oliva, E. et al. 2012, Proc. SPIE, 84462N

Links

The agendas of Council and STC meet ings can befound on the ESO web pages: http://www.eso.org/ public/about-eso/committees /

Telescopes and Instrumentation Pasquini L. et a l., Paranal Instrumentation Programme

Figure 2. PlannedParanal instrumentationin 2019. One new instru-ment in integration, fourin design and construc-tion and one in Phase Aare also planned at thistime (see Table 1).

Table 1. Proposeddevelopment plan forthe Paranal instrumen-tation programme.One year of Phase A isexpected to be carriedout, and the overallduration is typically esti-mated as six to sevenyears. Delivery in thelast column refers tostart of integration inParanal for instrumentsor to the end of the inte-gration for infrastructureprojects (such as the

AOF and VLTI).

Year

2012

2013

2014

2015

2016

2017

2018

2019

2020

Phase A

CUBESCRIRES upgrade

Letter of interestNTT

New I (NTT?)

New II

New III

New IV

New V

New VI

Design & Construction

ERIS

MOONSCRIRES upgrade

4MOST

CUBES (?)

New I (NTT?)

New II

New III

New IV

New V

Delivery

KMOS VIMOS upgrade

MUSESPHERE

VISIR upgradePRIMA astrometryGRAVITYLFC for HARPS

AOFMATISSE

ESPRESSO VLTI

CRIRES upgrade

CUBES(?)MOONS

ERIS4MOST

New I (NTT?)

UT1 (Antu)

CRIRESKMOSFORS2

UT2 (Kueyen)

UVESMOONS X-shooter

UT3 (Melipal)

VIMOSSPHERE VISIR/CUBES

VISTA

4MOST

VLTI

AmberGRAVITY MATISSEPRIMA

UT4 (Yepun)

MUSEHAWK-IERIS AOF

ESPRESSO

http://www.eso.org/public/about-eso/committeeshttp://www.eso.org/public/about-eso/committeeshttp://www.eso.org/public/about-eso/committeeshttp://www.eso.org/public/about-eso/committees


7/72


Telescopes and Instrumentation

Revisiting the Impact of Atmospheric Refraction on

VIMOS-MOS Observations: Beyond the Two-hour Angle

Rule

Rubén Sánchez-Janssen1

Fernando Selman2

Steffen Mieske2

Paul Bristow2

Peter Hammersley2

Michael Hilker2

Marina Rejkuba2

Burkhard Wolff2

1 NRC Herzberg Institute of Astrophysics,Canada

2 ESO

Multi-object spectroscopic (MOS)observations with VIMOS have tradi-tionally been limited to a narrowtwo-hour range from the meridian tominimise slit losses caused by atmos-pheric dispersion and differentialrefraction. We revisit the impact ofthese effects on the quality of VIMOS-MOS spectra through extensive simu-lations of slit losses. We show thatMOS observations can be effectivelyextended to plus/minus three hoursfrom the meridian for elds with zenithangles smaller than 20 degrees atculmination — provided a nonstandardrotator offset angle of 0 degrees isused. The increase in target observabil-ity will enhance the efciency of opera-tions, and hasten the completion ofprogrammes — a particularly relevantaspect for the forthcoming spectro-scopic public surveys with VIMOS.

Atmospheric refraction in VIMOS-MOSobservations

VIMOS (Le Fevre et al., 1998) is a wideeld-of-view (four elds of 7 by 8 arc-minutes) instrument with imaging, integraleld, and multi-object spectroscopiccapabilities mounted at the Nasmyth Bfocus of the Very Large Telescope (VLT)Unit Telescope 3. The instrument oper-ates in the optical wavelength range(360–1000 nm), and is equipped with sixsets of grisms, six sets of broadbandlters, plus three additional lter setsspecically designed to be used in com-bination with the grisms to block the

second-order spectra. During the last fewyears the instrument performance hasbeen signicantly enhanced (seeHammersley et al., 2012; 2013): changing

the detectors to red-sensitive, low-fringing CCDs; replacing the HR-bluegrism set with higher throughput volumephase holographic grisms; introducingan active exure compensation system;redesigning the focusing mechanismand mask cabinet; and introducing a newpre-image-less MOS mode (Bristow etal., 2013). All these improvements havemade VIMOS a much more stable instru-ment, and have extended its lifetime toprepare it for the start of the spectro-scopic public surveys for which ESO hasrecently issued a call.

Further work to improve the operationalefciency of the instrument includes the

present study, which has, as its maingoal, to revisit the need for restrictedobservability of targets only within plusand minus two hours from the meridian in

the MOS mode — the two-hour anglerule. VIMOS is not equipped with atmos-pheric dispersion compensators (ADCs),and MOS observations are carried outusing multi-slit masks (see Figure 1). Asa result, atmospheric dispersion (causedby the wavelength variation of the indexof refraction of air) and eld differentialrefraction (resulting from airmass varia-tions across the eld of view [FoV]) intro-

duce a wavelength-dependent uxreduction, due to slit losses, that cannotbe corrected. Unfortunately, eld rotationfurther prevents the alignment of all slits

Figure 1. Example of a VIMOS nding chart for MOSobservations. Each quadrant is 7 by 8 arcminutes,and they are separated by two arcminute gaps.

Allocated sli ts are overplotted in blue. The blank

areas (upper right of each nding char t) are maskedout proposal information.


8/72


9/72


a certain level of data compression.Following the previous work by Cuby etal. (1998), we set the tolerance level forlosses/distortions at 20%. In Figure 5 weshow the declination–hour-angle pairs(colour-coded according to slit orienta-tion) for which the median spectral d istor-tion (top row) or median ux loss (bottom

row) across the VIMOS FoV remain belowthis tolerance value during a one hourlong integration. It is evident that for eldsculminating at small zenith distances the

larger distortions and ux losses occur atlarger HAs. Both the amount of losses/ distortions, and the dependence on dec-lination, increase for bluer wavelengths. On the other hand, for the east–west(PA = 90 degree) orientation, we see thatat xed HA there is a very strong depend-

ence on declination, but the behaviourattens towards redder wavelengths. Theminimum of the loss/distortion distribu-tions slightly decreases and moves

towards southern declinations at largerHAs. For any given grism, there is verylittle dependence on HA (except for ex-treme declinations). Finally, the depend-ence on declination of losses/distortionsincreases towards bluer wavelengths.

Beyond the two-hour angle ruleExtracting simple rules from a problemwith such a high dimensionality requires

Figure 2. Output simu-lated spectra for ninedifferent slit positionsacross the VIMOS FoV.

These are the result of aone hour long integra-tion (–3 < HA < –2) on aδ = 0 degree eld usingthe LR-red grism for aninput at spectrum. Ineach panel we show thespectra for two differentslit orientations, as wellas the correspondingrelative ux loss (f ) andspectral distortion (Δ).


10/72


with targets at δ ~> −5 or δ ~


11/72


range. Figure 6 shows the new airmassconstraint limits for MOS observationblocks. They have been signicantlyrelaxed for elds culminating at smallzenith distances, thus increasing targetobservability. This will enhance theefciency of operations, and speed upthe completion of programmes — aparticularly relevant aspect for the forth-coming spectroscopic public surveyswith VIMOS. These recommendations forMOS observations have already been inplace since September 2013. To denethe optimal slit position angle for any spe-cic target declination and instrumentsetup, we refer the users to the summaryplots in the slit losses report at the VIMOS news section1.

References

Bristow, P. et al. 2012, The Messenger, 148, 13Cuby, J.-G. et al. 1998, Proc. SPIE, 3355, 36Hammersley, P. et al. 2012, The Messenger, 142, 8Hammersley, P. et al. 2013, The Messenger, 151, 2Le Fèvre, O. et al. 1998, Proc. SPIE, 3355, 8

Links

1 Report on VIMOS slit losses: http://www.eso.org/

sci/facilities/paranal/instruments/vimos/doc/ rsjvimosslitlossessept2013.pdf

Figure 6. VIMOS airmass constrain ts for MO Sobserving blocks. The two different shadedareas correspond to the limits for elds thatcan be observed with slits having north–south

orientations at meridian (purple), or east–westorientations (green). The generating formulaefor these curves are shown w ith the samecolour coding.

Figure 5. Circles show thedeclination–hour-anglepairs (colour-coded accord-ing to slit orientation —orange for north–south andgreen for east–west) forwhich the median spectraldistortion (top row) ormedian ux loss (bottomrow) across the VIMOS FoVremain below 20% during aone hour long integration,for all the VIMOS grisms.

–90 –80 –70

1.1

1

1.2

1.3

1.4

1.5

1.6

M a x i m u m a

l l o w e d a i r m a s s

1.7

1.8

1.9

Stay in the shaded area when specifying your airmass constraint

VIMOS ai rmass constraints for MOS OBs

airmass ≤ 1/ (0.64280 cos δ – 0.41668 sin δ )airmass ≤ 1/ (0.78726 cos δ – 0.41668 sin δ )

2

2.1

2.2

2.3

2.4

–60 –50 –40

Target decl inat ion (δ )

–30 –20 0 10 20 30–10

http://www.eso.org/sci/facilities/paranal/instruments/vimos/doc/rsjvimosslitlossessept2013.pdfhttp://www.eso.org/sci/facilities/paranal/instruments/vimos/doc/rsjvimosslitlossessept2013.pdfhttp://www.eso.org/sci/facilities/paranal/instruments/vimos/doc/rsjvimosslitlossessept2013.pdfhttp://www.eso.org/sci/facilities/paranal/instruments/vimos/doc/rsjvimosslitlossessept2013.pdfhttp://www.eso.org/sci/facilities/paranal/instruments/vimos/doc/rsjvimosslitlossessept2013.pdfhttp://www.eso.org/sci/facilities/paranal/instruments/vimos/doc/rsjvimosslitlossessept2013.pdf


12/72


devices from e2v of high (but not perfect)cosmetic quality. In addition to the32 CCDs making up the science array,OmegaCAM also contains four auxiliaryCCDs around the edges of the eld thatare used for autoguiding and imageanalysis. Since both guiding and imageanalysis are performed on the instrumentside, the telescope “only” tracks, butdoes so very well. Outside a zenith-centriccircle of about 10 degrees diameter,image quality remains acceptable for upto ~ 2–3 minutes without guiding.

Most data are taken in the ve Sloan-likebands u, g, r , i and z , and a narrow-band Hα lter provided by the VPHAS+consortium. Service mode operations forOmegaCAM started on 15 October 2011. The median full width at half maximum(FWHM) of OmegaCAM images, asmeasured during the rst half year ofoperation, was about 0.80 arcseconds in i -band, and 0.95 arcseconds in g-band(including the instrumental resolutionof 0.4 arcseconds). In the rst two yearsof operations, the sky was clear or photo-metric 77% of the time.

Three public surveys are being executedat OmegaCAM (see Figure 3): the Kilo-Degree Survey (KiDS; 1500 squaredegrees), ATLAS (4500 square degrees),and the VST Photometric Hα Survey(VPHAS+; 2000 square degrees). Fortheir detailed setup and science goals,see the public survey web pages2 and

Steffen Mieske1 Dietrich Baade1 Magda Arnaboldi1 Giovanni Carraro1 Danuta Dobrzycka1 Armin Gabasch1 Philippe Gitton1 Nicolas Haddad1 Michael Hilker1 Ronald Holzloehner1 Valentin D. Ivanov1 Sebastien Morel1 Mark Neeser1 Loethe Noethe1

Ricardo Parra1 Andres Parraguez1 Monika Petr-Gotzens1

Andrew Rakich1

Marina Rejkuba1 Miguel Riquelme1 Fernando Selman1 Ricardo Schmutzer1 Thomas Szeifert1

1 ESO

The science operations process of the VLT Survey Telescope (VST) camera,OmegaCAM, is described. OmegaCAMis a 267-megapixel CCD camera imag-ing a 1 × 1 degree eld of view with apixel scale of 0.21 arcseconds. It beganoperations in October 2011. The tele-scope and camera provide a surveyspeed that is ve times greater than thenow-decommissioned Wide FieldImager on the MPG/ESO 2.2-metre tele-scope at La Silla. OmegaCAM is cur-rently used for three public surveys,

guaranteed time observations for theOmegaCAM and VST consortia, andChilean programmes. The execution ofOmegaCAM observations, real-timequality control and the calibration planare outlined.

General description of the facility

The VST resulted from a collaborationbetween the Italian National Institute of Astrophysics ( INAF) under the PrincipalInvestigator (PI) Massimo Capaccioli andESO (see Capaccioli & Schipani [2011]for a description). OmegaCAM1 was builtin a collaboration between ESO andthe OmegaCAM consortium (PI, KonradKuijken) with contributions from theNetherlands, Germany and Italy and isdescribed by Kuijken et al. (2002) andKuijken (2011). With OmegaCAM ESOfullled its mandate from the ESO Councilto provide an optical wide-eld imagingcapability at Paranal Observatory.

OmegaCAM is the wide-eld imager forthe Cassegrain focus of the VST onParanal. The VST is a 2.6-metre modiedRitchey–Chretien alt-azimuth telescope(F/5.5) designed specically for wide-eldimaging, to exploit the good image qualityat the Paranal Observatory. OmegaCAMobserves from 330–950 nm within acorrected eld of view of 1 × 1 degree,four times the size of the full Moon.OmegaCAM samples the VST eld ofview with a 32-CCD, 16k × 16k detectormosaic with 0.21-arcsecond pixel scale(Figure 2). The CCDs are thinned, blue-sensitive, three-edge buttable CCD44-82


OmegaCAM Science Operations

Figure 1. Left: The VST on the VLT platform. Right:OmegaCAM (the yellow volume) mounted on the VST.

E S O / I N A F - V S T / O m e g a C A M / G .

L o m b a r d i


13/72


the dedicated articles in this Messenger edition. Releases of data from thesesurveys are available through Phase 33. The execution of these three sur veystakes up 50–60% of the available servicemode time. The remaining 40–50% isshared between guaranteed time obser-vations (GTO) for the OmegaCAM and VST consortia, Chilean programmes, andcalibration observations. The respectiveshares may evolve over time as a functionof programme completion rates.

The survey speed of OmegaCAM isabout ve times higher than the WideField Imager (WFI) at the MPG/ESO2.2-metre telescope in La Silla (Baade etal., 1999). The good image qualit y andhigh blue sensitivity provide a uniquewindow in observational parameter space.Furthermore, it complements very well

the near-infrared survey telescope VISTAat Paranal (Emerson & Sutherland, 2010),which is operated at the neighbouringconsole in the VLT control room. VSTand VISTA together cover the entire near-ultraviolet to near-infrared range 0.33–2.35 μm: VST from 0.33–0.95 μm, and VISTA from 0.85–2.35 μm, with an over-lap at the z -band that is used by bothfacilities.

Execution of observations

All observations on the VST are carr iedout in service mode. The two surveyswith short integration times, VPHAS+ and ATLAS, observe most of the time in openloop (no guiding, no image analysis). Dueto the high cadence of ~ 1–2 minutesbetween images taken at different posi-tions on the sky in these two surveys,there is no time to acquire image analysisstars and perform active optics correc-tion at each new position. A new correc-tion is enforced after, at most, half an

hour in such open loop observations. The KiDS survey spends more time ona single pointing due to the sciencerequirement of providing a deep andaccurate weak lensing map. KiDS obser-vations are, therefore, performed withguiding and closed loop image analysis.

The telescope and instrument are oper-ated at night by one telescope operator,without a night-time astronomer. Theshort-term scheduling of observations isdone by a program called the Observing Tool (OT; see also Bierwir th et al. [2010]). The basic observation unit containingthe full description of the observationsequence (acquisition and science expo-sures), information about the targetand the required observing constraintsnecessary to achieve the scientic objec-tive, is called an Observation Block (OB). At any given time, the OT lters out allOBs that are not observable in the cur-rent conditions (due to seeing, airmass,Moon illumination, sky transparency),and then ranks the observable OBs tomatch, as well as possible, the observingconditions requested by the users, while

Figure 2. Left: An image showing the 32 (4 × 8) CCDarray of OmegaCAM and the four auxiliary CCDs usedfor guiding and wavefront analysis. Right: OmegaCAMdome at illustrating individual chip sensitivities, whichhave a root mean square chip-to-chip variation of± 6%. The 32 science CCDs cover 92% of the1 × 1 degree eld of view with only small inter-chipgaps. Dither patterns of a few arcminutes total ampli-tude are used to image the full eld of view.

Figure 4. Distinguished visitors at the VSTOmegaCAM console: Chilean president SebastianPiñera and his wife. Also present in the picture isParanal staff astronomer Fernando Selman.

Figure 3. The on-sky footpr int of the threepublic surveys executed with OmegaCAM: KiDS(1500 square degrees), ATLAS (4500 square degrees),

VPHAS+ (2000 square degrees).


14/72


suggests quality control grades for eachOB to the night-time operator (see Fig-ure 6). Specically, the script calculates,for each frame across the eld of view,the mean FWHM, the mean ellipticity andthe IQ variation, which is the variationin the FWHM in the centre of the eld vs.the FWHM at the edge of the eld ofview. The script also measures the num-ber of single CCD images affected byellipticities greater than 0.2. These num-bers are then appropriately averagedacross an OB, or concatenation of OBs,and the script suggests to the operatorthe zero-level quality control (QC0) grades(fully/almost/not within constraints). Thegeneral QC0 acceptance criteria for anOB are: average seeing ≤ 1.1 × requestedseeing; average ellipticity ≤ 0.15; and IQvariation ≤ 25%.

In addition to these criteria for a singleOB, the QC0 script includes a number ofnested criteria regarding single images,concatenations of OBs, special cases likeGTO and agreements with consortiaabout deviations from the general QC0acceptance criteria per OB. The scriptalso contains a number of warning agswhich highlight image quality and cali-bration plan issues to the operator. Thefull set of these criteria is quite complexand could not realistically be trackedmanually by an operator during the night. Therefore the QC0 script is a crucial partof OmegaCAM science operations. Itallows fully reproducible quality controlaccording to criteria agreed betweenESO and the users that is as independentas possible from variations in humanhabits. An example output is indicated inFigure 6, left panel.

taking into account relative programmepriorities (see Figure 5 for an exampleof the OT ranking). Users prepare theirOBs using the P2PP4 software whichincludes the option to dene relationsbetween OBs (groups, concatenationsand time series). Those higher levelconstraints are also included in the rank-ing made by the OT.

It is worth noting that around full Moonthere is a scarcity of observable OBsdue to the bright sky background, whichmost of the science cases for opticalimaging cannot tolerate. Also, only a fewprogrammes accept the presence of thinclouds. The combination of those twoissues has led to a comparably high frac-tion of 8–10% idle time at the VST. ESOis taking measures to improve the bafingof the VST which, in turn, will reduce theeffect of scattered moonlight on the sci-ence images. More detail is provided inthe section on challenges and outlook.

Real-time quality control during night-time observations

The high cadence of observations andlarge number of 32 CCDs (267 mega-

pixels) requires reliable automated qualitycontrol. With one person operating boththe telescope and the instrument at night,it is not possible to measure manuallyimage quality parameters across the eldfor all those images, which can amountto several hundred per night. Real-timequality control is performed by using theoutput log of the OmegaCAM data reduc-tion pipeline, which was assembled byESO from algorithmic modules providedby the OmegaCAM consortium. Thereis a 1–2 minute delay between the com-pletion of an image and the availabilityof the pipeline output log. This allowsnear real-time assessment of data quali ty,enables fast decisions to be taken tore-adjust the input parameters for the OTranking engine, and hence optimise theobservation plan for the next hour(s). The pipeline provides measurements ofmean point spread function (PSF) FWHMand ellipticities for detected sources ineach single chip.

The core of the automated qualit y contro lis a dedicated script with some 800 linesof code which reads in the pipeline out-put log, calculates the specic parame-ters that are used for quality controlassessment (called QC0 at ESO), and


Figure 5. Examplescreenshot of ObservingBlock ranking as per-formed by the Observ-ing Tool.

Figure 6. Left: Example output of the real-time qual-ity control script used for OmegaCAM. Each line cor-responds to one image. The grouped sets of images

correspond to one OB or a concatenation of OBs.Right: PSF anisotropy distribution for a single imageof average FWHM 0.65 arcseconds, taken in EarlyScience.

Mieske S. et al., OmegaCAM Science Operations


15/72


The set of QC0 criteria appl ied toOmegaCAM night-time observationsensures that we full our mission state-ment of providing data of excellent imagequality to the users. This very strictadherence to a combination of criterialeads to a slightly larger rate of OB repe-titions than for other instruments: forOmegaCAM, about 20% of time in ser-vice mode is spent on repeating OBs(note that all data, whether in or out ofconstraints, is immediately transferred tothe ESO archive). The typical fractionfor other non-adaptive optics instrumentsat Paranal is 10–15%. Paranal instru-ments using adaptive optics, and VLTinterferometer instruments, have agreater than 20% fraction of time spenton repeating observations since they veryoften push out to the instrumental andatmospheric limits. For OmegaCAM, thebalance between strict QC0 and quickobserving progress is constantly reviewedby the operations team in consultationwith the survey consortia.

Calibration plan

The cal ibration plan of OmegaCAMensures continuous monitoring of thesystem throughput in the u, g, r , i and z -bands. Sky ats are taken in two to

three lters during each clear eveningtwilight. Equatorial Landolt photometricstandard star elds in u to z -bands areobserved at the beginning and in the

middle of the night. Furthermore, a shortobservation close to the southern celes-tial pole is performed three times pernight in a segmented lter with simultane-ous u, g, r and i coverage. These highairmass observations guarantee continu-ous monitoring of the extinction, comple-menting the low-airmass Landolt eldobservations. Other lters like Johnson B and V or Hα are calibrated with Landoltstandards only if science data are takenin these lters. Based on dedicated ditherobservations of the Landolt elds in allchips and under photometric conditions,the OmegaCAM consortium has builtup secondary standard star catalogues(currently with more than 315 000 stars)in the key bands for the full 1 × 1 degreeeld of view.

Daytime calibrations consist of daily biasand dome ats, and weekly gain/linearitymeasurements. Quality control duringthe day focuses on monitoring the chipsensitivities, bias levels, dark currentlevels, readout noise, gain and linearity,at lamp intensities, twilight at levels,magnitude zeropoints and image quality(FWHM and ellipticity). A descriptionof the OmegaCAM health checks5 andthe resulting scores and plots6 is main-tained. This follows the classic ESOapproach of scored health checks main-tained by the QC group in Garching(Hanuschik et al., 2008). An example ofthe detector-monitoring health check isshown in Figure 7.

Challenges and outlook

VST operations have non-negligible over-heads, impacting the completion progress

of the public surveys, which were plannedbefore the telescope and camera per for-mance parameters were measured. Mostnotably, the time used to bring and keep

the telescope focal plane to its bestshape is higher than originally expected.In the rst two years of OmegaCAMoperations, about 30% of the availablescience time was used for acquisition,mainly image analysis. This can be com-pared to 10% for VISTA. Related to this,OmegaCAM has a larger fraction of d irectinteraction with the system for the tele-scope operator when compared to VISTA.

Work is ongoing, in collaboration with theinstrument and telescope consortia, toimprove the control procedures of the VST main and secondar y mirrors, and tooptimise image analysis algorithms onthe OmegaCAM side. The aim is to movecloser towards fully automated opera-tions, including automatic acquisition ofguide and image analysis stars.

Another area of improvement is the skyconcentration effect that produces rota-tionally asymmetric features of 5% ampli-tude in the sky ats (see the OmegaCAMuser manual7 ), and reection/scat teringfeatures in images close to the Moon(within a few tens of degrees). Calibratingout the sky at variations requires greatcare in data reduction (see the consor-tium report on sky concentration correc-tion7 ). Additional bafing for the VSTwill be tested at the end of 2013; this willimprove the reproducibility of the skyconcentrations in sky ats, and permitobservations closer to the Moon, thusreducing idle time.

Survey progress and rst results

In Figure 8 the progress of the VST publicsurveys during the rst two years of

operations is shown. In total, about 1000hours of observing time per year havebeen spent on executing public surveyOBs within user constraints.

Figure 7. An example of an O megaCAM heal th-check scoring performed by the QC group in

ESO Garching. The left panel shows the dome atlevel averaged over all 32 detectors, while the middlepanel shows the median level for each detector. Thepanel on the right plots the score status of eachdetector (labelled with their names: ESO_CCD_#65to ESO_CCD_#96) as a function of time. If the domeat level falls below or exceeds levels dened for eachdetector, then the square at that date will tur n red.


16/72


total KiDS/VIKING area of 1500 squaredegrees is about 40, with > 10 above z = 6.5.

References

Baade, D. et al. 1999, The Messenger, 95, 15Bierwirth, T. et al. 2010, SPIE, 7737E, 19Capaccioli, M. & Schipani, P. 2011, The Messenger,

146, 2Emerson, J. P. & Sutherland, W. 2010, SPIE,

7737E, 4Hanuschik, R. W. et al. 2008, SPIE, 7016E, 22Kuijken, K. 2002, The Messenger, 110, 15Kuijken, K. 2011, The Messenger, 146, 8

Venemans, B. et al. 2013, ApJ, 779, 24Wright, N. J. et al. 2013, MNRAS, in press,

arXiv:1309.4086

Links

1 ESO OmegaCAM webpage: http://www.eso.org/ sci/facilities/paranal/instruments/omegacam

2 Setup and goals of the VST public surveys:http://www.eso.org/sci/observing/PublicSurveys/ sciencePublicSurveys.html

3 ESO Phase 3 page:http://www.eso.org/sci/observing/phase3.html

4 Phase 2 P2PP software: https://www.eso.org/sci/ observing/phase2/P2PP3.html

5 OmegaCAM quality control pages: http://www.eso.org/observing/dfo/quality/OMEGACAM/qc/qc1.html

6 Health check scores and pl ots: http://www.eso.org/observing/dfo/quality/OMEGACAM/common/ score_overview.html

7 OmegaCAM user manual and consortium repor t onsky concentration correction: http://www.eso.org/ sci/facilities/paranal/instruments/omegacam/doc /

In Figure 9, a few rst science resultsfrom VPHAS+ and KiDS, kindly providedby the survey teams, are illustrated. In the VPHAS+ example ( left), an ionised nebulasurrounding the extreme red supergiant,W26, begins to be resolved (from Wrightet al., 2013). As the only known exampleof a compact ionised nebula around ared supergiant, this represents a unique

opportunity to study the mass loss of redsupergiants, using the tools of nebulaastrophysics. The right panel shows oneexample of the nine-band u to Ks photo-metry of KiDS (OmegaCAM) and VIKING(VISTA Kilo-Degree Infrared GalaxySurvey) being used to hunt for high-red-shift quasi-stellar object (QSO) candi-dates which drop-out in the KiDS i -band(z > 5.7) and VIKING z -band ( z > 6.5);from Venemans et al. (2013). The histo-gram shows the redshift distribution of allpublished quasars at z > 5.7 (in grey).Currently, more than 50 quasars at z > 5.7have been discovered in various surveys. The red histogram bars show the redshiftdistribution of the quasars found in thecombined KiDS/VIKING survey thus far. The number of quasars expected in the


Figure 8. Cumulative number of hours, since star t ofoperations, spent on completed OBs for the three

VST OmegaCAM public su rveys . About 50–60% ofthe available service mode time at VST is spent onpublic surveys. The rest is shared between GTO andChilean programmes.

Figure 9. Left: A composite g-Hα- i image of thedense, very massive cluster Westerlund 1, taken aspart of VPHAS+, is shown. This image allows thestudy of the ionised nebula surrounding the ex tremered supergiant, W26 (Wright et al., 2013). Thezoomed part of the image (centre), shown in orange,is Hα only. Right: High-redshift QSOs detected bycombining KiDS (OmegaCAM) and VIKING (VISTA)data (Venemans et al., 2013), indicated as red in thehistogram. See text for more details.

Right ascension06.2 4

2 . 0

3 8 . 0

3 4 . 0

– 4 5 : 5 0 : 3 0 . 0

D e c l i n a t i o n

2 6 . 0

05.8 05.4 16:47:05.0

Mieske S. et al., OmegaCAM Science Operations

http://www.eso.org/sci/facilities/paranal/instruments/omegacamhttp://www.eso.org/sci/facilities/paranal/instruments/omegacamhttp://www.eso.org/sci/observing/PublicSurveys/sciencePublicSurveys.htmlhttp://www.eso.org/sci/observing/PublicSurveys/sciencePublicSurveys.htmlhttp://www.eso.org/sci/observing/phase3.htmlhttps://www.eso.org/sci/observing/phase2/P2PP3.htmlhttps://www.eso.org/sci/observing/phase2/P2PP3.htmlhttp://www.eso.org/observing/dfo/quality/OMEGACAM/qc/qc1.htmlhttp://www.eso.org/observing/dfo/quality/OMEGACAM/qc/qc1.htmlhttp://www.eso.org/observing/dfo/quality/OMEGACAM/common/score_overview.htmlhttp://www.eso.org/observing/dfo/quality/OMEGACAM/common/score_overview.htmlhttp://www.eso.org/observing/dfo/quality/OMEGACAM/common/score_overview.htmlhttp://www.eso.org/sci/facilities/paranal/instruments/omegacam/dochttp://www.eso.org/sci/facilities/paranal/instruments/omegacam/dochttp://www.eso.org/sci/facilities/paranal/instruments/omegacam/dochttp://www.eso.org/sci/facilities/paranal/instruments/omegacam/dochttp://www.eso.org/observing/dfo/quality/OMEGACAM/common/score_overview.htmlhttp://www.eso.org/observing/dfo/quality/OMEGACAM/common/score_overview.htmlhttp://www.eso.org/observing/dfo/quality/OMEGACAM/common/score_overview.htmlhttp://www.eso.org/observing/dfo/quality/OMEGACAM/qc/qc1.htmlhttp://www.eso.org/observing/dfo/quality/OMEGACAM/qc/qc1.htmlhttps://www.eso.org/sci/observing/phase2/P2PP3.htmlhttps://www.eso.org/sci/observing/phase2/P2PP3.htmlhttp://www.eso.org/sci/observing/phase3.htmlhttp://www.eso.org/sci/observing/PublicSurveys/sciencePublicSurveys.htmlhttp://www.eso.org/sci/observing/PublicSurveys/sciencePublicSurveys.htmlhttp://www.eso.org/sci/facilities/paranal/instruments/omegacamhttp://www.eso.org/sci/facilities/paranal/instruments/omegacam


17/72


E S O P

u b l i c S u r v e y s

The two VLT imaging survey telescopes: the Visi-

ble and Infrared Survey Telescope for Astronomy(VISTA, upper) and the optical VLT Survey Tele-

scope (VST, lower).

E S O / G . H ü d e p o h l ( a t a c a m a p h o t o . c o m )

E S O / G . H ü d e p o h l ( a t a c a m a p h o t o . c o m )

http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/http://www.atacamaphoto.com/


18/72


19/72


OmegaCAM (Kuijken, 2011). An overviewof OmegaCAM’s scientic operations isgiven in Mieske et al. in this issue (p. 12). The VST surveys are starting their thirdyear of operation. KiDS is the largest sur-vey, requiring 3225 hrs to completion,or 39% of the time, followed by 15% for VST ATLAS and 11% for VPHAS+. Thetime allocated to Chilean programmes(10%), and Guaranteed Time Observa-tions (GTO) of the OmegaCAM consor-tium (15%) and the Italian National Insti-tute of Astrophysics (INAF; 10%) make upfor a sizable fraction of the available tele-scope time, which has an importantimpact on the speed of completion. Fig-ure 2 is a summary pie chart of the timeallocation of the VST surveys as a per-centage of the total available VST time forthe public surveys, Chilean and GTO pro-grammes.

The rst spectroscopic public survey, theGaia–ESO survey, started operation on

1 January 2012 on FLAMES at the VLTUnit Telescope 2 (Kueyen) on Paranal. The data acquis ition for this survey is car-ried out in visitor mode and the timeallocation of the survey entails 60 nightseach year, for an overall assignmentover four years initially, with another yearpending a review of the survey progressand delivered data. Thus far 105 nightshave been allocated to the Gaia–ESOsurvey. The PESSTO survey started oper-ation on 1 April 2012, on EFOSC andSOFI at the New Technology Telescope(NTT) on La Silla. The data acquisition forthis survey is also carried out in visitormode. The time allocation for the surveyincludes 90 nights each year, with anallocation of 60/30 nights in odd andeven periods respectively, for an overallassignment over four years, also withanother year pending a successfulreview. Thus far 120 nights have beenallocated to this project.

Progress and estimated completion timefor imaging surveys

An integral par t of the approval of publ icsurvey projects is the review of their sur-vey management plans (SMPs), whichoutline the plans for telescope time allo-cation and observing constraints overthe years. Additional information on qual-ity control and pipeline data reduction,survey resource allocation for the surveyexecution (full-time equivalent [FTE] allo-cation), the timeline for the delivery andthe description of the data products for

publication in the SAF are all part of theSMPs. Hence the SMPs have becomethe benchmark that is used to computethe progress of the public surveys andthey represent the basis for estimatingtheir completion time. In service mode,the basic observation unit is the observa-tion block (OB) and the time charged tothe programme is accounted for in termsof the number of successfully completedOBs. This includes the shutter open time(exposure time) and the relevant over-heads provided in the execution timereporting module, which is part of theobservation preparation software (P2PP3).

Successfully completed OBs are exe-cuted observations that full the requestedobserving constraints according to strin-gent quality control (QC) criteria that areexplained extensively elsewhere1. Furtherinformation on the VST/OmegaCAM QCprocess, which was designed followingthe QC for VISTA/ VIRCAM observations,is given in the article by Mieske et al. (p.12). In the following, the fraction of thecompletion for a survey is computed as afraction of the total execution time inhours for the completed OBs normalisedby the total time in hours requested inthe approved SMPs. The cumulative dia-grams for the percentage of completionare shown in Figures 3 and 4 for the VISTA and VST surveys, respectively.

VISTA — The six VISTA surveys are pro-gressing at a similar pace. As for anynew telescope the start of VISTA opera-tions required some adjustments. After

Chilean

GTOOmegaCAM

GTO INAF

VST ATLAS VPHAS+

KiDS

Figure 2. Pie chart showing the time allocation aspercentage of the total available time for the VST.

Figure 3. Graph of the percentage completion forthe VISTA surveys as a function of date.

Figure 4. Graph of the percentage completion forthe VST surveys as a function of date.


20/72


asked to issue informed recommenda-tions on the continuation of surveyprogrammes, or their termination, shouldthey consider any of them not scienti-cally competitive at the time of the review.

Publication and download of sciencedata products from ESO public surveys

The ESO policies in place to managethe public survey projects monitor thedelivery of data products for ingestionand publication via the SAF. Additionalallocation of telescope time is conditionalon the submission of data products viaPhase 3. Phase 3 concludes the processstarted with the submission of the letterof intent, followed by Phase 1 (proposalpreparation and submission) and thepreparation and submission of OBs forobservations in service mode, i.e.,Phase 2. As a result of Phase 3, the com-munity can access and download thedata products from the SAF and is ableto carry out independent science projects in addition to those targeted by thesurvey teams (c.f., Arnaboldi et al., 2011).

major technical interventions in 2010(camera shim installation and horizontalre-centring) and 2011 (primary andsecondary mirror recoating and extendedrecovery), operational procedures wereadopted to increase the speed of execu-tion and reduce the number of repeatedOBs. The current completion rate for the VVV and VHS surveys is more than 67 %,while the percentage of completion is inthe range 52% to 42 % for the other sur-veys (UltraVISTA, VIKING, VIDEO and VMC).

VST — For the VST surveys, the per-centages of completion are 66% for VST ATLAS, 22 % for KiDS and 38% for VPHAS+. Of the three VST surveys, KiDShas the tightest requirements in termsof Moon illumination (dark time) and see-ing constraints. It is also, by far, the larg-est survey on the VST and thus, even ifit uses a comparable fraction of time peryear to the other surveys, its overall com-pletion is much smaller. Furthermore theright ascension/declination distributionof the target elds overlaps with those ofapproved GTO projects. Strategies arebeing implemented to mitigate the com-petition, and speed up the data acquisi-tion for KiDS, and also improve the over-all observation progress on the VST.For more details see Mieske et al. (p. 12).

Taking account of the above percentagesof completion for each survey and the

start of operation of each survey tele-scope, and assuming the same observa-tion progress as previously, we can eval-uate the date of completion for the VST

and VISTA surveys. Figures 5 and 6 showthe expected completion time in yearsand the year of completion for the nineimaging surveys. We expect the VVV, VHS and VST ATLAS to be completed by2015, with the other surveys coming tocompletion in the following years, withKiDS completed in 2021. For VHS, twonumbers are shown in Figure 5: the com-pletion in 2015 is based on the 3400hours requested in the SMP, but sincethe overheads were not known at thetime of writing the SMP, this survey willactually need about two years longer tocover the entire southern hemisphere, asshown by the light blue bar. It is impor-tant to point out that these projections donot automatically translate into telescopetime allocated to these surveys. Theseestimates are upper limits since, as onesurvey nishes, the others may progressfaster, which is not explicitly taken intoaccount in the simple extrapolationabove. The legacy value and the scienticexcellence of the survey programmesare considered by the public survey pan-els organised by ESO and these com-pletion dates are presented at major peerreviews. The public survey panels are

ESO Public Surveys

Table 1. Summary of VISTA and VSTpublic survey products in the ESOscience archive (Status: 25 October2013).

Survey

VHS

VIKING

VV V

VMC

Ultra-VISTA

VIDEO

ATLAS

VPHAS+

KIDS

Bands

YJHKs ZYJHKs ZYJHKs

YJKsYJHKsYJHKs

ugriz ugri, Hα

ugri

Sky coverage*

(sq.deg)

42102355643.61.81.8

234137556

Data volume

(GB)

8511288

2877268624

3015747701

Figure 5. Expected completion time, in years and byyear of completion, for the VISTA surveys. For VHSthe time to completion for the whole area coverageis indicated in light blue. T = 0 refers to the star t ofscientic operations, i.e., April 2010.

Figure 6. Expected completion time, in years andby year of completion, for the VST surveys. T = 0refers to the start of scientic operations, i.e.,October 2011.

Arnaboldi M. et al., The ESO Publ ic Surveys

* The quoted sky coverage is the tota lgeometric area of images, whichnormally differs from the nominal sur-vey area.


21/72


The year 2013 has been very importantfor Phase 3 activities, as all eleven ESOpublic surveys submitted and published

their data products via the SAF. The mile-stones for Phase 3 were the second VISTA submission for images and sourcelists, and the rst submission for cata-

logues. The rst data release of the VSTsurveys was announced in September2013. The spectroscopic public surveysare actively going through the processof content validation and it is p lanned thatthey will reach publication via the SAFby December 2013. Thus far, a total vol-ume of 16 TB of data products —

images, weight maps, source lists andcatalogues — is now available and full ysearchable via dedicated query inter-faces. In Table 1 we provide an overview

of the data volume, wavelength andsky coverage of the data releases fromthe imaging surveys. Further informationand detailed descriptions of the datareleases from the ESO public surveys areavailable2.

Public survey data are published throughthe ESO archive interfaces conjointly withother products such as the stream forthe ultraviolet and visual echelle spectro-graph (UVES) data that results from thein-house generation of science dataproducts. All Phase 3 data productscomply with the established standard forESO science data products, therebyguaranteeing uniformity in terms of dataformat and characterisation across theESO archive.

Figure 7 illustrates the current sky cover-age of the ESO survey products in twoprojections. More than 4500 squaredegrees in the NIR bands and 2400square degrees in the optical bands havebeen covered by data products, whichare now accessible via the query inter-faces of the SAF.

Merit parameters for ESO public surveysare the number of refereed publicationsby ESO survey teams and archive users,the number of press releases and thecumulative download of data productsfrom the ESO archive. There are now71 refereed publications from the surveyteams with a signicant increase in thenumber of refereed publications (+200%)since November 2012, including fromfour archive users, i.e., researchers whoare not members of the survey teams. The contribution by Wegg & Gerhard(p. 54) is an example of exciting scienticresults achieved using ESO archive dataproducts (in this case from the VVV sur-vey). There also are more than ten pressreleases based on VISTA data and morethan four press releases for the VST.

The parameters on the data downloadby the community also demonstrate astrong interest. The cumulative downloadfrom the SAF since December 2011amounts to more than 6.8 TB of dataproducts and ~ 27 000 les. In Figures 8

and 9 these numbers are differentiatedper survey project and data product type,respectively. The community is clearlyeager to access the data, with the largest

Figure 7. Sky coverage of ESO public sur vey prod-ucts is shown in two projections. Upper: Full sky(Hammer–Aitoff projection); lower: Southern hemi-sphere (stereographic projection).


22/72


volume download coming from VVV,UltraVISTA and KiDS; see Figure 8. Thelargest volume download for products isfor the source lists, followed by the tileimages. We believe that catalogues willrepresent very valuable assets, as theyare the highest level products for thesurveys. In this respect, we are workinghard to reach a critical data volume soon,with the ingestion of the VIKING, VVVand VMC catalogues so that the commu-nity can benet even more from the jointeffort of ESO and the survey teams.

References

Arnaboldi, M. et al. 1998, The Messenger, 93, 30 Arnaboldi, M. et al. 2007, The Messenger, 127, 28 Arnaboldi, M. et al. 2011, The Messenger, 144, 17Emerson, J., McPherson, A. & Sutherland, W. 2006,

The Messenger, 126, 41Kuijken, K. 2011, The Messenger, 146, 8

Links

1 Quality control criteria:http://www.eso.org/sci/observing/phase2/ SMGuidelines/ConstraintsSet.VIRCAM.html

2 Phase 3 data releases: http://www.eso.org/sci/

observing/phase3/data_releases.html

ESO Public Surveys

Figure 9. Number ofles downloaded for thedifferent data producttypes from the ESO SAFfor the public imagingsurveys.

Figure 8. Data volumedownload for the imag-ing public surveys.

Arnaboldi M. et al., The ESO Publ ic Surveys

The 2.6-metre VLT Survey Telescop e (VST ) is show nin its enclosure on Cerro Paranal. In the backgroundare the nearby VLT Unit Telescopes 3 (Melipal, to theright) and 4 (Yepun, left).

http://www.eso.org/sci/observing/phase2/SMGuidelines/ConstraintsSet.VIRCAM.htmlhttp://www.eso.org/sci/observing/phase2/SMGuidelines/ConstraintsSet.VIRCAM.htmlhttp://www.eso.org/sci/observing/phase3/data_releases.htmlhttp://www.eso.org/sci/observing/phase3/data_releases.htmlhttp://www.eso.org/sci/observing/phase3/data_releases.htmlhttp://www.eso.org/sci/observing/phase3/data_releases.htmlhttp://www.eso.org/sci/observing/phase2/SMGuidelines/ConstraintsSet.VIRCAM.htmlhttp://www.eso.org/sci/observing/phase2/SMGuidelines/ConstraintsSet.VIRCAM.html


23/72


both at the VISTA Science Archive and atESO. Further details about the VMC survey1 are given in Cioni et al. (2011).

Stellar populations

One of the main goals of the VMC survey isthe identication and characterisation ofthe mixture of stellar populations that havemade up the Magellanic system over time.

The star formation history of eld stars, thephysical parameters of stellar clusters, thelinks between these and the structure anddynamical processes are all embedded in the

VMC data. Ex tracting a comprehensive pic-ture of the system represents our major chal-lenge, but fortunately we have access tosophisticated tools with which to do the job.In Rubele et al. (2012), we demonstrated thatby using two colour–magnitude diagrams(CMDs) simultaneously, and a grid of modelsat various ages and metallicities, we couldderive spatially resolved SFHs where system-atic errors in the star formation rate and age–metallicity relations are reduced by a factor oftwo, relative to previous work, after account-ing for the geometry of the galaxy. In our studywe independently derive the mean extinctionand distance modulus for twelve subsectionsof the original tiles.

In Figure 1 we show the CMD of a tile in theSMC including the Milky Way (MW) globularcluster 47 Tuc, highlighting the complexity ofthe SFH analysis in decomposing the differentstellar populations. Using custom-derivedpoint spread function photometry, we canpush the sensitivit y of the VMC data to highlycrowded regions. Together with the wide areacovered by VMC we will be able to investigatenot only substructures in the LMC and SMC,but also streams at tached to the 47 Tuc clus-ter, for example, as well as detecting themembers of hundreds of stellar clusters in theMagellanic system waiting to be characterised.

The reddening map of the 30 Doradus eld

Dust causes uncertainties in the measure-ments of the SFH and the structure of galax-ies. Red clump stars (0.8–2 M and 1–10 Gyrold) are useful probes of interstellar reddeningbecause of their large number and relativelyxed luminosity. Red clump stars belongingto the tile LMC 6_6 are selected from theirlocation in the ( J–Ks) vs. Ks CMD. Then, theamount of total reddening (along the l ine ofsight and within the LMC) in terms of colourexcess is obtained for each of ~ 150 000 starswith respect to its intrinsic colour. The latteris derived accordingly from stellar evolution

Maria-Rosa L. Cioni1,2

Peter Anders3

Gemma Bagheri1

Kenji Bekki4

Gisella Clementini5

Jim Emerson6

Chris J. Evans7

Bi-Qing For4

Richard de Grijs8

Brad Gibson9

Léo Girardi10

Martin A. T. Groenewegen11

Roald Guandalini12

Marco Gullieuszik10

Valentin D. Ivanov13

Devika Kamath12

Marcella Marconi14

Jean-Baptiste Marquette15

Brent Miszalski16

Ben Moore17

Maria Ida Moretti14

Tatiana Muraveva5

Ralf Napiwotzki1

Joana M. Oliveira18

Andrés E. Piatti19

Vincenzo Ripepi14

Krista Romita20

Stefano Rubele10

Richard Sturm21

Ben Tatton18

Jacco Th. van Loon18

Mark I. Wilkinson22

Peter R. Wood23

Simone Zaggia10

1 University of Hertfordshi re, United Kingdom2 Leibniz-Institut für Astrophysik Potsdam,

Germany3 National Astronomical Observatory of

China, China4 ICRAR, University of Western Australia,

Australia5 INAF, Osservatorio Astronomico di Bologna,

Italy6 Queen Mary University London, United

Kingdom7 Astronomy Technology Centre, Edinburgh,

United Kingdom8 Peking University, China9 University of Central Lancashire, United

Kingdom10 INAF, Osservatorio Astronomico di Padova,

Italy11 Royal Observatory of Belgium, Belgium12 Institute of Astronomy, KU Leuven, Belgium13 ESO14 INAF, Osservatorio Astronomico di

Capodimonte, Italy15 Institut d’Astrophysique de Paris, France16 South African Astronomical Observatory,

Cape Town, South Africa

17 University of Zurich, Switzerland18 Lennard-Jones Laboratories, Keele Univer-

sity, United Kingdom19 Observatorio Astronómico, Universidad

National de Córdoba, Argentina20 University of Florida, USA 21 Max-Planck-Institut für extraterrestrische

Physik, Germany22 University of Leicester, United Kingdom23 Australian National University, Australia

The VISTA near-infrared YJKs survey of the

Magellanic Clouds system (VMC) has

entered its core phase: about 50 % of the

observations across the Large and Small

Magellanic Clouds (LMC, SMC), the

Magellanic Bridge and Stream have already

been secured and the data are processed

and analysed regularly. The initial ana lyses,

concentrated on the rst two completed tiles

in the LMC (including 30 Doradus and the

South Ecliptic Pole), show the superior qual-

ity of the data. The photometric depth of

the VMC survey allows the derivation of the

star formation history (SFH) with unprece-

dented quality compared to previous wide-

area surveys, while reddening maps of high

angular resolution are constructed using

red clump stars. The multi-epoch Ks-band

data reveal tight period–luminosity relations

for variable stars and permit the measure-

ment of accurate proper motions of the stel-

lar populations. The VMC survey continues

to acquire data that will address many issues

in the eld of star and galaxy evolution.

The VMC survey

The VMC survey observations are obta inedwith the infrared camera VIRCAM mounted on

VISTA and reach sta rs down to a limitingmagnitude of ~ 22 (5σ Vega) in the YJKs lters.

The VMC strategy involves repeated observa-tions of tiles across the Magellanic system,where one tile covers approximately uniformlyan area of ~ 1.5 square degrees in a givenband with three epochs at Y and J, and 12epochs at Ks spread over a time range of oneyear or longer. Individual Ks epochs refereach to exposure times of 750 s and reach alimiting magnitude of ~ 19 for sources withphotometric errors < 0.1 mag. The VMC dataare acquired under homogeneous sky condi-tions, since observations take place in servicemode, and their average quality correspondsto a full width at half maximum < 1 arcsecond.

The VISTA astrometry, which is based on2MASS, results in positional accuracies within25 milliarcseconds (mas) across a tile. The

VMC data are reduced with the VISTA DataFlow System (VDFS) pipeline and are archived

ESO Public Surveys

The VMC ESO Public Survey


24/72


models, accounting for variations wi th ageand metallicity. Extinction is subsequentlyconverted into hydrogen gas column density.Compared to reddening maps producedusing the same method at optical wave-lengths, the near-infrared VMC data are moresensitive to higher extinction. Compared toH I observations we der ive that, on average,half of the stars lie in front of the H I columnand hydrogen becomes molecular in thedustiest clouds; the transition begins atN H I ≅ 4 × 10

21 cm–2. Figure 2 shows thelocation of molecular clouds superposed onthe distribution of hydrogen column densityinferred from the VMC data. There is overallagreement with maps of dust emission at24 μm and 70 μm (see Tatton et al., 2013).Reddening maps will be created for other tilesin the VMC survey allowing red clump starsto be de-reddened; these results will be usedin calculating the three-dimensional (3D) struc-ture of the Magellanic system.

Variab le stars

The other main goal of the VMC survey isthe measurement of the 3D structure of theMagellanic system. Classical Cepheids areprimary distance indicators and in the near-infrared obey period–luminosity (PL) relations

that are less affected by reddening, chemicalcomposition and nonlinearity than those atoptical wavelengths, resulting in smaller intrin-sic dispersion. First results for classicalCepheids in the tiles LMC 6_6 (Figure 3, left)and 8_8 have been presented in Ripepi et al.(2012). The identication of the variab les isderived from the EROS-2 and OGLE-III cata-logues and their VMC Ks light curves are verywell sampled, with at least 12 epochs, andhigh precision, with typical errors of 0.01 mag,or better, for individual phase points. The Ks mag of the faintest Cepheids in the LMC,which are mostly rst over tone pulsators, wasmeasured for the rst time thanks to the

VMC observing strategy. Photometr y for thebrightest fundamental mode Cepheids (peri-ods > 23 days), exceeding the linearit y regimeof VMC data, are taken from the literature. Thedispersion of the PL relations is ~ 0.07 mag.

Anomalous Cepheids (1.3–2.1 M and

[Fe/H] ≈ –1.7 dex) also play an impor tant roleboth as distance indicators and stellar popula-tion tracers. The VMC survey has alreadyobserved many of the anomalous Cepheidsdiscovered by the OGLE project in the LMC.

These stars obey a tight PL relation in theKs-band with a dispersion of 0.10 mag (Fig-ure 3, right) that is shown for the rst time inRipepi et al. (2013).

Cepheids (< 200 Myr old) are mainly concen-trated towards the bar and in a nor thwest

spiral arm of the LMC as well as in the centralregion of the SMC. Eclipsing binaries com-posed of main sequence stars trace a simi lardistribution, but with clustering mainly occur-ring in regions of recent star formation. Onthe other hand, RR Lyrae variable stars(> 10 Gyr old) are smoothly distributed andlikely trace the haloes of the galaxies. Thesestars also follow a PL relation that is tightin the Ks-band. The VMC properties and thestrategy to measure distances and infer thesystem 3D geometry of different age compo-nents from the variable stars is described inMoretti et al. (2013).

The magnitude of the brightest VMC objects(10 < Ks < 12), which may be saturated in theircentral regions, is well recovered by the VDFSpipeline by integrating the ux in the outerparts. Most of these sources are asymptoticgiant branch (AGB) stars. By tting spectralenergy distributions, created from the combi-nation of VMC data and data at other wave-lengths, with dust radiative transfer models, itis possible to derive mass-loss rates, luminosi-ties and spectral classications that offerstrong constraints on AGB evolutionary mod-els (Gullieusz ik et al., 2012). These variablestars obey PL relations that may also be usefulas distance and structure indicators.

The proper motion of the LMC

The astrometric accuracy and the photometricsensitivity of observations made with VISTAare of sufcient quality to select a large sam-ple of targets and measure their proper motion.

The proper motion of the LMC is measuredfrom the combination of 2MASS and VMCdata that span a time range of ~ 10 years andfrom VMC data alone across a time baselineof ~ 1 year (Cioni et al., 2013b). Dif ferent typesof LMC stars (e.g., red giant branch, red clumpand main sequence stars, as well as variablestars) are selected from their location in the( J–Ks) vs. Ks CMD, and from lists of knownobjects, where MW foreground stars and back-ground galaxies are also easily distinguished(Figure 4, left). The proper motion of ~ 40 000LMC stars in the tile, with respect to ~ 8000background galaxies, is μαcos(δ ) = +2.20 ±0.06 mas yr –1 and μδ = 1.70 ± 0.06 mas yr

–1. This value is in exce llent agreement wi th previ-ous ground-based measurements but ourstatistical uncer tainties are a factor of threesmaller and are directly comparable to uncer-tainties derived wi th the Hubble Space Tele-scope. The error budget is at present domi-nated by systematic uncertainties (a few masyr –1), but these will decrease due to theimproved reduction of the VISTA data and theincrease in the time baseline.

ESO Public Surveys

Figure 1. Colour–magnitude diagram of stellarsources in tile SMC 5_2. All sources are shown ingrey; stars belonging to the SMC cluster NGC 121are indicated in blue; the eld population of theSMC is indicated with yellow contours; and the starsof the Milky Way cluster 47 Tuc are shown with red

contours.

Figure 2. Hydrogen column density map inferredfrom VMC data in tile LMC 6_6 identifying regionswhere N H I > 8 × 10

21 cm–2. Crosses representmolecular clouds catalogued in the literature andellipses highlight those with measured properties.

Cioni M.-R. L. et al., The VMC ESO Public Survey


25/72


formation of the Bridge and the existence ofstripped stars. The VMC survey has a high leg-acy value and represents the sole counterpartin the Ks-band to current and future ground-based (STEP at the VST, SkyMapper, SMASHat the Blanco 4-metre, Large Synoptic Survey

Telescope [LSST]) and space-based imagingmissions (e.g., Gaia and Euclid) targeting orincluding the Magellanic system. It also pro-vides a wealth of targets for wide-eld spectro-scopic follow-up investigations, e.g., with the

Apache Point Observatory Galactic EvolutionExperiment (APOGEE)-South, High Efciencyand Resolution Multi-Element Spectrograph(HERMES) at the Anglo-Australian Observatory,4-metre Multi-Object Spectroscopic Telescope(4MOST) and Multi-Object Optical and Near-infrared Spectrograph (MOONS) at ESO.

Refer

messenger no154

Documents