Astronomical Science

The VISTA Near-infrared YJKs Public Survey of the Magellanic Clouds System (VMC)

Maria-Rosa Cioni1, 2 13 UK Astronomy Technology Centre, debate as to whether it is also of tidal Gisella Clementini3 United Kingdom ­origin from the interaction of the Large Leo Girardi4 14 Centre for Astrophysics, University of Magellanic Cloud (LMC) with the Small Roald Guandalini1 Central Lancshire, United Kingdom Magellanic Cloud (SMC), or produced by Marco Gullieuszik5 15 University of Cambridge, Institute of ram pressure between the two galaxies. Brent Miszalski1 Astronomy, United Kingdom Maria-Ida Moretti6 16 ESO The VMC1 survey is acquiring near-infra­ Vincenzo Ripepi7 17 UPMC, University of Paris, Institut red (NIR) data of unprecedented sensi­ Stefano Rubele4 d’Astrophysique de Paris, France tivity in the Magellanic system that is of Gemma Bagheri1 18 CNRS, Institut d’Astrophysique de immense value for the astronomical com­ Kenji Bekki8 Paris, France munity. The survey represents the only Nick Cross9 19 LERMA, Observatoire de Paris, France NIR counterpart to existing optical sur­ Erwin de Blok10 20 University of Zurich, Institute of Theo­ veys and for the large number of unclas­ Richard de Grijs11 retical Physics, Switzerland sified ob­­jects observed with the Spitzer Jim Emerson12 21 University of Exeter, School of Physics, Space Telescope in the mid-infrared. The Chris Evans13 United Kingdom VMC will cover the bulk of the Magellanic Brad Gibson14 22 University of Keele, School of Physical system as opposed to the tiny regions Eduardo Gonzales-Solares15 and Geographical Sciences, United sampled by the Hubble Space Telescope, Martin Groenewegen5 Kingdom and the limited area covered by most of Mike Irwin15 23 University of Leicester, United King­ the other ground-based observations at Valentin Ivanov16 dom the same sensitivity. Jim Lewis15 24 Mount Stromlo Observatory, RSSA, Marcella Marconi7 Australia Jean-Baptiste Marquette17, 18 Chiara Mastropietro19 Ben Moore20 The VISTA public survey project VMC Ralf Napiwotzki1 targets the Large Magellanic Cloud, Tim Naylor21 the Small Magellanic Cloud, the Bridge Joana Oliveira22 and two fields in the Stream. The VMC Mike Read9 survey is a uniform and homogeneous 9 Eckhard Sutorius survey in the Y, J and Ks near-infrared Jacco van Loon22 filters. The main goals are the determi- Mark Wilkinson23 nation of the star formation history and Peter Wood24 the three-dimensional structure of the Magellanic system. The survey is there- Figure 1. VMC logo fore designed to reach stars as faint 1 University of Hertfordshire, Physics as the oldest main sequence turn-off Astronomy and Mathematics, United point and to constrain the mean magni- Current open questions about the for­ Kingdom tude of pulsating variable stars such mation and evolution of the Magellanic 2 University Observatory Munich, as RR Lyrae and Cepheids. We provide Clouds will be addressed by the VMC ­Germany a brief overview of the survey strategy survey, such as: How have the SFHs of 3 INAF, Osservatorio Astronomico di and first science results. Further details the LMC and SMC been influenced by Bologna, Italy are given in Cioni et al. (2011). interaction? Does the geometry of the 4 INAF, Osservatorio Astronomico di Magellanic system depend on age and Padova, Italy metallicity? Why is there a significant 5 Royal Observatory of Belgium, Belgium Introduction ­difference in structure between the gas 6 University of Bologna, Department of and stars in the SMC? Astronomy, Italy The Magellanic Clouds offer an excellent 7 INAF, Osservatorio Astronomico di laboratory for near-field cosmology. They Capodimonte, Italy represent the closest prototype for stud­ Observations and data reduction 8 ICRAR, University of Western Australia, ies of interacting galaxies and their low Australia metallicity and high gas content provide VISTA 2 (Emerson et al. 2006) is the larg­ 9 University of Edinburgh, Institute for information about galaxies at an early est wide-field NIR imaging telescope Astronomy, United Kingdom stage of evolution. The Magellanic Clouds designed to perform survey operations. 10 University of Cape Town, South Africa have experienced an extended star for­ The performance during commissioning 11 Peking University, Kavli Institute for mation history (SFH) and their dynamical is presented by Emerson et al. (2010), Astronomy and Astrophysics, China interaction may be responsible for the while science verification programmes 12 Queen Mary, University of London, formation of the Magellanic Bridge. The are summarised in Arnaboldi et al. (2010). United Kingdom origin of the Magellanic Stream is under

The Messenger 144 – June 2011 25 Astronomical Science Cioni M.-R.. et al., The VISTA Near-infrared YJKs Public Survey

The VMC survey parameters are listed observed in the LMC. One field covers sky levels. Tile catalogues are produced in Table 1. VMC uses three filters, Y, J the famous 30 Doradus region, one field following the application of a nebulos- and Ks for the analysis of colour–colour corresponds to the South Ecliptic Pole ity filter in order to remove diffuse varying diagrams; the wider colour spacing in (SEP) region, there are a pair of fields background on scales of 3 arcseconds

Y–Ks provides a good characterisation of located in the northern outer part of the or larger (Irwin 2010). The average VMC the sub-giant branch population for LMC disc while the remaining two fields parameters from all single-tile images deriving the SFH. Monitoring of variable are located towards the Bridge. The are given in Table 3 where the magnitude sources is performed in the Ks-band, ­progress of the VMC survey observations limit is at 5σ. This set comprises ob­­­ where variable stars obey a clear period– up to 4 March 2011 is provided in Table 2. servations obtained up to the end of magnitude relation. The VMC survey area November 2010 but does not distinguish is shown in Figure 2. It covers the area The VISTA raw images are reduced by between crowded and uncrowded fields to a limiting B magnitude of 25 mag arc­ the VISTA Data Flow System (VDFS) for which the VMC has different observ­ sec–2 for both galaxies and encompasses ­pipeline at the Cambridge Astronomical ing requirements. the major features traced by the distribu­ ­Survey Unit (CASU3). Standard reduc­- tion of stars and gas. tion steps are performed and quality con­ Astrometry is based on the positions of trol parameters are calculated for both the many 2MASS sources within each Here we focus on observations obtained monitoring and evaluating observing con­ detector. The median astrometric root- during the dry-run period (November ditions retrospectively. A tile image is pro­ mean-square is 80 milliarcseconds (mas) 2009–March 2010) when VISTA was duced by combining different pawprint and is dominated by uncertainties on tested and survey operations were still images adjusted for astrometric and pho­ 2MASS coordinates. Residual systematic being defined. Six VMC fields were tometric distortions as well as different distortions across the VISTA field of view are present at the 25 mas level. The pho­ tometric calibration relies on the observa­ Table 1. VMC survey parameters. tion of stars from the 2MASS catalogue with magnitudes in the range 12–14 in all

Filter Y J Ks Filter Y J Ks bands. The calibration in the Y-band is Central λ (μm) 1.02 1.25 2.15 Exposure time per epoch (s) 800 800 750 possible where the extinction is not too Bandwidth (μm) 0.10 0.18 0.30 No. of epochs 3 3 12 high, i.e. EB–V < 1.5 (Hodgkin et al. 2009). DIT (s) 20 10 5 Total exposure time (s) 2400 2400 9000 Figure 3 shows the behaviour of VISTA No. of DITs 4 8 15 Sensitivity per epoch (Vega) 21.3 20.8 18.9 photometric uncertainties in a VMC field. No. of exposures 1 1 1 S/N per epoch 5.7 5.9 2.9 Uncertainties are reduced by about 50% Micro-stepping 1 1 1 Total sensitivity (Vega) 21.9 21.4 20.3 compared to those for individual tiles No. of jitters 5 5 5 Total S/N 10 10 10 and will reduce further for deep tiles. A Pawprints in tile 6 6 6 Saturation (Vega) 12.9 12.7 11.4 morphological classification flag is also Pixel size (arcsec) 0.339 0.339 0.339 Area (deg2) 184 184 184 included in the catalogue. System FWHM 0.51 0.51 0.51 No. of tiles 110 110 110 The data reduced by the VDFS pipeline are ingested into the VISTA Science Description VMC LMC SMC Bridge Stream Table 2. VMC survey Archive4 (VSA). At present these data are No. of tiles 110 68 27 13 2 progress up to 4 March reduced with the version 1.0 of the 2011. (Items in blue refer No. of epochs 1980 1224 486 234 36 to the complete survey.) ­pipeline and include all VMC data ob­­ No. of Y epochs 330 204 81 39 6 served until end of November 2011. At No. of J epochs 330 204 81 39 6 the VSA the data are archived to produce

No. of Ks epochs 1320 816 324 156 24 standardised data products: individual Observed Y epochs 59 37 8 8 6 passband frame association and source Observed J epochs 58.5 38.5 6.5 7.5 6 association to provide multi-colour, multi-

Observed Ks epochs 98.5 75 5 4.5 14 epoch source lists; cross-associations No. of observed epochs 216 150.5 19.5 20 26 with external catalogues; and deeper Completion in Y 17.9 % 18.1 % 9.9% 20.5% 100 % stacking. The position and magnitude of Completion in J 17.7 % 18.9 % 8% 19.2 % 100 % each source in a given table refers to the

Completion in Ks 7.5 % 9.2% 1.5 % 2.9% 58.3% astrometrically and photometrically cali­ Total completion 10.9 % 12.3 % 4% 8.5% 72.2% brated measurements using the parame­ ters specified in the image headers. In addition the magnitude of the brightest Filter FWHM Ellipticity Zero-point Mag. Limit Table 3. VMC survey tile quality up to stars (Ks < 12 mag.) are recovered for sat­ Y 1.03 (0.14) 0.06 (0.01) 23.44 (0.10) 21.00 (0.49) 30 November 2010. uration effects (Irwin 2009). Figure 4 J 1.01 (0.13) 0.06 (0.01) 23.68 (0.14) 20.55 (0.44) shows the Ks magnitude difference for Ks 0.92 (0.10) 0.05 (0.01) 23.00 (0.19) 19.25 (0.30) stars in a VMC field compared to 2MASS before and after saturation correction.

26 The Messenger 144 – June 2011 5.5h 5h 4.5h 4h 3.5h 3h 2.5h 2h 1.5h 1h

–58°

M

EA

–56° R

ST –60°

–58° –62° )

00 –60°

20 –64° (J δ

–62° –66°

–64° –68°

7.5h 7h 6.5h 6h 5.5h 5h 4.5h 4h3.5h 3h 2.5h 2h 1.5h 1h 0.5h 0h 23.5h23h 22.5h α (J2000)

Figure 2. (Top) Magellanic system area tiled for VMC Figure 3. (Bottom left) Photometric uncertainties in Figure 4. (Bottom right) Magnitude difference for observations superimposed on the distribution of the VMC data for stacked pawprints (dashed, dotted stars in common between VMC and 2MASS before

H i gas (McClure-Griffiths et al. 2009). Tiles are col­ and continuous lines in the Ks-, J- and Y-bands (bottom) and after (middle) recovering the magnitude our coded as follows: purple: completed; cyan: respectively) and tiles (red, green and blue circles in of VMC stars approaching the saturation limit; the observations in progress; green: to be observed. the Ks-, J- and Y-bands respectively) in one field are adjustment applied is shown at top. shown.

2 CORR 15 C VM S

10 –K 0

5 CORR

NO 0 S

K –2 200 13 14 15 16 2 g) ma CORR

S

–3 0 –K 0 (1 2M S g K ma σ –2 K 100 S J 2 CORR Y NO S 0 –K 2M S K –2 0 14 16 18 20 10 12 14 mag KS

The Messenger 144 – June 2011 27 Astronomical Science Cioni M.-R.. et al., The VISTA Near-infrared YJKs Public Survey

Figure 5. Complete- the red giant branch (RGB) beginning at 1 ness results for the ~2 mag below the red clump, the South Ecliptic Pole 0.8 (SEP) field are shown for ap­proximately circular region described 0.6 single epoch (in green) by the highest concentration of stars. and deep stacked 0.4 The RGB continues beyond the red images (in red). clump at brighter magnitudes describing 0.2 Y-band a narrow structure bending to red col­ 0 ours. The abrupt change in source den­

s 1 sity at the tip of the RGB marks the 0.8 ­transition to brighter asymptotic giant enes

et 0.6 branch (AGB) stars. The broad vertical 0.4 distribution of stars below the RGB is

Compl 0.2 J-band populated by Milky Way (MW) stars that 0 are easily distinguished from LMC stars. Cepheid and supergiant stars occupy the 1 region of the diagram to the bright and 0.8 blue side of the RGB, while RR Lyrae 0.6 stars are somewhat fainter than the red 0.4 clump and lie more or less parallel to the 0.2 sub-giant branch. KS-band 0 14 16 18 20 22 24 MAG The Ant diagram

12 10 10 Figure 7 shows the observed colour- colour diagram, and the corresponding 14 CMDs, of the VMC data in the SEP field. The distribution of sources in the colour– 16 colour diagram resembles the body of an 15 15 Y S S ant, with different parts distinguished: 18 K K – Gaster (–0.1 < (Y–J) < 0.3 and –0.3 <

(J–Ks) < 0.3), shown in red, the lower 20 part of the ant body corresponds to 20 20 the location of MS stars in the LMC, 22 with the youngest being at the bluest extremity. –0.5 0 0.5 1 0 1 0 12 – Mesosoma (0.2 < (Y–J) < 0.4 and Y–J J–KS Y–KS 0.4 < (J–Ks) < 0.75), shown in blue, the middle part of the ant body corresponds Figure 6. Colour–magnitude diagrams of VMC added and recovered stars. The results to the main locus of helium-burning sources in the SEP field. of this process are shown in Figure 5. giants in the LMC. Most of them are in the red clump. The completeness of the stellar photome­ – Petiole (0.15 < (Y–J) < 0.3 and 0.3 < try as a function of location across the Colour–magnitude diagrams (J–Ks) < 0.4), shown in green, is a small Magellanic sytem and position in the col­ thin extension of the mesosoma at its our–magnitude diagrams is estimated via Figure 6 shows the colour–magnitude dia­ red side, and is mainly caused by fore­ the usual procedure of adding artificial grams (CMDs) of VMC stellar and proba­ ground bright stars in the MW stars of known magnitudes and positions ble stellar sources in the SEP field. These – Head (0.4 < (Y–J) < 0.55 and 0.65 < to the images, then comparing them in data were extracted from the VSA. The (J–Ks) < 0.85), shown in magenta. Its the derived photometric catalogues. For exposure times per band correspond to main blob is defined by foreground this work tile images are created and 2400 s in Y, 2800 s in J and 9400 s in Ks. low-mass stars in the Milky Way. regions equivalent to 1/12 of the area are – Upper antenna (0.42 < (Y–J) < 0.6 and selected for performing PSF photometry. The distribution of stars in the CMDs 0.85 < (J–Ks) < 1.1), shown in cyan, Then, artificial stars are added randomly shows different stellar populations. The being formed by the more luminous at a mutual distance of > 30 pixels, so bluest conic structure, bending to red RGB stars in the LMC close to the tip of as not to increase the typical crowding of colours at bright magnitudes, is formed the RGB, extending up to (J–Ks) = 1 mag. the images. Artificial stars spanning by main sequence (MS) stars of increas­ – Lower antenna ((Y–J) > 0.55 and smaller bins are later grouped together to ing mass with increasing brightness. 0.6 < (J–Ks) < 0.85), shown in yellow, estimate the number ratio between The MS joins, via the sub-giant branch, corresponding to the (Y–J) redward

28 The Messenger 144 – June 2011 4.5 Ant diagram 1.5 1.5 Ant diagram Upper antenna 4 Upper antenna Lower antenna 1 1 Lower antenna 3.5 s

Head g) S

K 3 0.5 0.5 Head g( J– Lo

J– K Mesosoma Mesosoma Petiole 2.5 Petiole 0 0 Gaster 2 Gaster –0.5 –0.5 1.5

–0.4 –0.2 0 0.2 0.4 0.6 0.8 1 1.2 1.4 0 0.5 1 Y–J Y–J

10 12 4.5

12 4 14

14 3.5

16

s 3 Y g) 16 K Log( 18 2.5 18

2 20 20

1.5

22 22 012 00.5 11.5 22.5

Y–Ks Y–Ks

Figure 7. Colour–colour (“Ant diagrams”) and colour– The simulated Ant diagram is also shown the K-band that is only weakly af­­fected magnitude (CM) diagrams of VMC sources in the in Figure 7, where colours now represent by evolutionary effects, the spread in stel­ SEP field are shown for observed magnitudes (upper and lower left) and simulated magnitudes (upper and the stellar surface gravity. lar mass within the instability strip and lower right). In the observed magnitude plots (left uncertainties in reddening corrections. column), the coloured points correspond to the mor­ Similarly, the Cepheid period–luminosity phological selection outlined in the text and labelled RR Lyrae stars and Cepheids relation in the K-band is much narrower in the upper plot. The mapping of stellar surface gravity to the simulated colour–colour and CM dia­ than the corresponding optical relations, grams is shown in the right column. Radially pulsating stars obey a period- and less affected by systematic uncer­ mean density relation that forms the tainties in reddening and metal content. basis of their use as standard candles

to measure the distance to the host sys­ In the context of the VMC survey, the Ks extension of foreground low-mass stars tem. In particular, RR Lyrae stars obey photometry is taken in time series mode

in the MW. a period–luminosity–metallicity relation in in order to obtain mean Ks magnitudes

The Messenger 144 – June 2011 29 Astronomical Science Cioni M.-R.. et al., The VISTA Near-infrared YJKs Public Survey

P = 0.601586 d; Epoch = 52581.80256 JD P = 3.87046 d; Epoch = 51566.70416 JD Figure 8. BEROS and Ks VMC light- α = 90.32399 deg; δ = –66.74809 deg α = 88.87615 deg; δ = –66.09933 deg curves for an RR Lyrae star (left) and 17.5 15.0 a Cepheid (right) in the SEP field. B15574 B6 104 18.0 15.2

18.5 15.4 ros ros e e B 19.0 B 15.6

19.5 15.8

20.0 16.0 –1.0 0.0 1.02.0 –1.0 0.0 1.02.0 17.0 13.8 K15574 K6 104

17.5 14.0 s s K K 18.0 14.2

18.5 14.4 –1.0 0.0 1.02.0 –1.0 0.0 1.02.0 Phase Phase

Figure 9. Colour-composite image of LMC PN MG 60 made from single (left) and stacked (right) exposures in Y

(blue), J (green) and Ks (red) pass­ bands.

for RR Lyrae stars and Cepheids over the are 117 RR Lyrae and 21 Cepheids in shallower observations (shown in green), whole Magellanic system. Their period– common between EROS-2 and VMC. confirming the soundness of our observ­ luminosity relations will be used to meas­ EROS-2 periodicities were confirmed by ing strategy, and allowing us to derive ure distances and construct a 3D map analysing their BEROS light-curves with mean magnitudes without using template of the stellar distribution. The identifica­ GRATIS (Clementini et al. 2000). Figure 8 light-curves. The average of individual tion, period and epoch of maximum light shows BEROS and VMC light-curves of the Ks measures corresponds to 18.02 +/– for these objects are taken from micro­ RR Lyrae star #15574 (P = 0.601586 0.12 mag. and 14.17 +/– 0.06 mag. for the lensing surveys (EROS, MACHO and days) and the Cepheid variable #6104 RR Lyrae star and Cepheid, respectively. OGLE) of the LMC and SMC while in the (P = 3.87046 days). For the RR Lyrae star Bridge we plan to use the VST data there is one point per night and the themselves (Cappellaro, 2005). ­magnitude is the weighted average of all Planetary nebulae available observations, while for the The location of the SEP field at the ­Cepheid two points, corresponding each Planetary Nebulae (PNe) are well known periphery of the LMC overlaps with to a different detector, are shown. Both for their role in the development of the EROS-2 (Tisserand et al., 2007). There light-curves are well sampled, including extragalactic standard candle planetary

30 The Messenger 144 – June 2011 11

13

15 S K 17

19

21

–0.5 0.0 0.5 1.01.5 2.0

J–KS Figure 10. Colour-composite image (2 by 2 arc­ minutes, shown right) and CMD (shown left) of the stellar cluster KMHK 1577. The isochrone corre­ sponds to an age of 0.63 Gyr and a metallicity of Z = 0.004. nebula luminosity function (PNLF). Dis­ mainly misclassified field stars, compact and 9400 s in Ks; triangles indicate ­stellar tances can be measured from the near- H ii regions or long-period variables. The sources and asterisks extended objects. universal bright end cut-off across all gal­ strongest emission lines in the NIR for axy types, but it remains difficult to PNe include: H i at 1.083 μm in Y-band, Acknowledgements explain how old stellar populations that Pa β in J, while Ks contains Br γ, multiple lack recent star formation episodes could H i and molecular H lines. Figure 9 2 The VMC survey is supported by ESO and the UK produce progenitors massive enough shows the impact of stacking individual STFC. This research has made use of Aladin, to power the high star luminosities popu­ VMC exposures to detect PNe. EROS-2 and 2MASS data. The UK’s VISTA Data lating the bright end. Flow System comprising the VISTA pipeline at CASU and the VISTA Science Archive at the Wide Field Astronomy Unit (WFAU) in Edinburgh has been cru­ Magellanic Cloud PNe are well positioned Stellar clusters cial in providing us with calibrated data products for to further advance our understanding of this paper. the PNLF. The deep NIR photometry pro­ Stellar clusters are among the primary vided by the VMC survey will allow new targets of the VMC survey. The detection References PNe to be detected and the enhanced of known stellar clusters will be exam- sen­sitivity to dust resulting from observa­ ined and new clusters will be searched Arnaboldi, M. et al. 2010, The Messenger, 134, 42 tion in the NIR will enable identification for. The analysis of stellar clusters will Cappellaro, E. 2005, The Messenger, 120, 13 Cioni, M.-R. L. et al. 2011, A&A, 527, A116 of contaminating compact H ii regions be centred on CMDs to estimate their Clementini, G. et al. 2000, AJ, 120, 2054 and symbiotic stars. The synoptic nature ages and metallicities as well as on the Emerson, J. et al. 2006, The Messenger, 126, 41 of the VMC survey, which allows varia­ comparison with results obtained from Emerson, J. et al. 2010, The Messenger, 139, 2 bility to be measured, will enable identifi­ optical surveys. Ages, masses and metal­ Hodgkin, S. T. et al. 2009, MNRAS, 394, 675 cation of pulsating Mira stars in the most licities of stellar clusters will allow us to Irwin, M. 2009, UKIRT Newsletter, 25, 15 Irwin, M. 2010, UKIRT Newsletter, 26, 14 obscured symbiotic stars, thus support­ discuss the SFH of the Magellanic Clouds Kontizas, M. et al. 1990, A&A, 84, 527 ing the cleaning of the PNe sample. and radial abundance gradients. McClure-Griffiths, N. et al. 2009, ApJS, 181, 398 Tisserand, P. et al. 2007, A&A, 469, 387 Within the first six VMC tiles in the LMC, Here we show results obtained for cluster a combination of optical imaging, OGLE KHMK 1577 (Kontizas et al. 1990) in the Links light-curves, the VMC NIR data and SEP field. The properties of this cluster SAGE mid-infrared observations, reveals are unknown and it was chosen because 1 VMC survey: http://star.herts.ac.uk/~mcioni/vmc/ 2 that only ~50% of the objects previously of its favourable location at the centre VISual and Infrared Telescope for Astronomy (VISTA): http://www.vista.ac.uk catalogued as PNe appear to be genuine. of a pawprint. The image and CMD of the 3 CASU: http://casu.ast.cam.ac.uk/surveys-projects/ These are characterised by the colours cluster are shown in Figure 10. The expo­ vista 4 VISTA Science Archive (VSA): http://horus.roe. 0.4 < (J–Ks) < 2 and 0.0 < (Y–J) < 0.5. sure time was 2400 s in Y, 2800 s in J The non-PNe identified in the sample are ac.uk/vsa/login.html

The Messenger 144 – June 2011 31