Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.arcetri.astro.it/science/starform/preprints/cesa_26.pdf Äàòà èçìåíåíèÿ: Mon Jul 27 15:47:43 2015 Äàòà èíäåêñèðîâàíèÿ: Sun Apr 10 09:50:40 2016 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: ï ï ï ï ï ï ï ï ï ï ï ï ï ï ï

Astronomy & Astrophysics manuscript no. 26199 July 23, 2015

c ESO 2015

Star and jet multiplicity in the high-mass star forming region IRAS 05137+3919
´ R. Cesaroni1 , F. Massi1 , C. Arcidiacono1,2 , M.T. Beltran1 , P. Persi3 , M. Tapia4 , S. Molinari3 , L. Testi5,1 , L. Busoni1 , 1 , K. Boutsia6 , S. Bisogni7,1 , D. McCarthy8 , and C. Kulesa8 A. Riccardi
1 2 3 4 5 6 7 8

INAF, Osservatorio Astrofisico di Arcetri, Largo E. Fermi 5, I-50125 Firenze, Italy cesa,fmassi,mbeltran,riccardi,lbusoni@arcetri.astro.it INAF, Osservatorio Astronomico di Bologna, Via Ranzani 1, I-40127 Bologna, Italy carmelo.arcidiacono@oabo.inaf.it INAF, Istituto di Astrofisica e Planetologia Spaziale, Via Fosso del Cavaliere 100, I-00133, Roma, Italy paolo.persi,sergio.molinari@iaps.inaf.it ´ ´ Instituto de Astronom´ , Universidad Nacional Autonoma de Mexico, Apdo. Postal 877, Ensenada, B. C., CP 22830, Mexico ia mt@astrosen.unam.mx European Southern Observatory, Karl-Schwarzschild-Str. 2, 85748, Garching, Germany e-mail: ltesti@eso.org INAF, Osservatorio Astronomico di Roma, Via di Frascati 33, I-00040 Monte Porzio Catone, Italy konstantina.boutsia@oa-roma.inaf.it ` Universita degli Studi di Firenze, Via Sansone 1, I-50019 Sesto Fiorentino, Italy e-mail: susanna@arcetri.inaf.it Steward Observatory, The University of Arizona, 933 N. Cherry Ave., Tucson, AZ-85721, U.S.A. dwmccarthy@gmail.com, ckulesa@as.arizona.edu

e-mail: e-mail: e-mail: e-mail: e-mail: e-mail:

Received date; accepted date
ABSTRACT

Context. We present a study of the complex high-mass star forming region IRAS 05137+3919 (also known as Mol8), where multiple jets and a rich stellar cluster have been described in previous works. Aims. Our goal is to determine the number of jets and shed light on their origin, and thus determine the nature of the young stars powering these jets. We also wish to analyse the stellar clusters by resolving the brightest group of stars. Methods. The star forming region was observed in various tracers and the results were complemented with ancillary archival data. The new data represent a substantial improvement over previous studies both in resolution and frequency coverage. In particular, adaptive optics provides us with an angular resolution of 80 mas in the near IR, while new mid- and far-IR data allow us to sample the peak of the spectral energy distribution and thus reliably estimate the bolometric luminosity. Results. Thanks to the near-IR continuum and millimetre line data we can determine the structure and velocity field of the bipolar jets and outflows in this star forming region. We also find that the stars are grouped into three clusters and the jets originate in the richest of these, whose luminosity is 2.4 â 104 L . Interestingly, our high-resolution near-IR images allow us to resolve one of the two brightest stars (A and B) of the cluster into a double source (A1+A2). Conclusions. We confirm that there are two jets and establish that they are powered by B-type stars belonging to cluster C1. On this basis and on morphological and kinematical arguments, we conclude that the less extended jet is almost perpendicular to the line of sight and that it originates in the brightest star of the cluster, while the more extended one appears to be associated with the more extincted, double source A1+A2. We propose that this is not a binary system, but a small bipolar reflection nebula at the root of the large-scale jet, outlining a still undetected circumstellar disk. The gas kinematics on a scale of 0.2 pc seems to support our hypothesis, because it appears to trace rotation about the axis of the associated jet.
Key words. Stars: early-type ­ Stars: formation ­ ISM: jets and outflows

1. Introduction
Observationally, the study of high-mass star formation is severely hindered by the difficulty of identifying the object of interest in a crowded region. This is basically the result of the large distances of OB-type stars (typically, several kpc) and the presence of numerous, lower mass stars in the same field. As a matter of fact, OB-type stars form in rich clusters, which makes it difficult to distinguish the phenomena associated with the highmass star(s) of interest from those associated with other cluster members. High angular resolution is thus crucial for this type of investigation, since the (projected) separation inside these rich stellar cluster may be 1 . Resolutions that high can be attained by IR and radio interferometers as well as by 8-m class

Send offprint requests to: R. Cesaroni, e-mail: cesa@arcetri.astro.it Based on observations carried out with the Large Binocular Telescope. The LBT is an international collaboration among institutions in the United States, Italy and Germany. LBT Corporation partners are: The University of Arizona on behalf of the Arizona university system; Istituto Nazionale di Astrofisica, Italy; LBT Beteiligungsgesellschaft, Germany, representing the Max-Planck Society, the Astrophysical Institute Potsdam, and Heidelberg University; The Ohio State University, and The Research Corporation, on behalf of The University of Notre Dame, University of Minnesota, and University of Virginia.


2

R. Cesaroni et al.: Star and jet multiplicity in the high-mass star forming region IRAS 05137+3919

telescopes operating at IR wavelengths, which are made possible through adaptive optics techniques. With this in mind, we have focused our attention on a massive star forming region, where evidence of multiple, young stellar objects (YSOs) was found. The corresponding counterpart in the IRAS PSC is IRAS 05137+3919, also known as Mol 8 from the catalogue of Molinari et al. (1996). These authors quote a kinematic distance of 10.8 kpc and a corresponding luminosity (estimated from the IRAS fluxes) of 5.6 â 104 L . Recently, Honma et al. (2011) have measured the parallax of the water +5 3 masers in this object, resulting in a distance of 11.6-2..8 kpc. The same authors also establish a lower limit of 8.3 kpc. This is consistent with a more recent parallax measurement by Reid +3 4 et al. (2014), who derive a distance of 7.7-1..8 kpc, implying an uncertainty of a factor 2 in luminosity. In the following, we adopt the minimum distance of 8.3 kpc estimated by Honma et al. (2011), with the caveat that higher values are possible. IRAS 05137+3919 does not satisfy the colour constraints established by Wood & Churchwell (1989) to identify (UC) HII regions in the IRAS PSC, because the colour index [60­12]=1.29 lies ­ although marginally ­ below the limit of 1.3. Indeed, despite the luminosity above 104 L , the source was not detected at 6 cm and 2 cm by Molinari et al. (1998), while only weak, compact continuum emission (0.33 mJy; Molinari et al. 2002) was measured at 3.6 cm. Whether this originates in an HII region or in a thermal jet still needs to be understood, but it appears likely that the 3.6 cm continuum emission is associated with one of the members of the embedded stellar cluster detected in the near IR by various authors (Ishii et al. 1998, 2002; Varricat et al. 2010) and studied by Faustini et al. (2009), Kumar et al. (2006), and Nikoghosyan & Azatyan (2014, 2015). The most likely candidates for ionising the radio source are two bright stars located at the centre of the cluster and already identified by Varricat et al. (2010). Interstingly, that cluster lies also close to the geometrical center of a bipolar outflow mapped in the 12 CO(1­0) line by Zhang et al. (2001, 2005). Varricat et al. (2010) imaged two bipolar jets in the H2 v=1­0 S(1) line, also centred on the stellar cluster and directed approximately NE­SW and NW­SE. Hereafter, we refer to these, respectively, as Jet 1 and Jet 2, following Varricat et al. (2010). The former is roughly oriented like the 12 CO outflow by Zhang et al. (2001, 2005) and the HCO+ (1­0) outflow mapped by Molinari et al. (2002) with the OVRO interferometer. The presence of shocked H2 emission is confirmed by the near-IR spectra obtained by Ishii et al. (2001), which refer to the NW­SE jet. Based on these previous findings, we decided to perform a multi-wavelength study of IRAS 05137+3919 with the main goal to relate the properties of the stellar cluster to those of the outflows/jets and hence identify the sources powering the flows and establish their nature. For these purposes, we obtained outflow maps a factor 4 better in angular resolution than those by Zhang et al. (2005) and imaged the stellar cluster with an 8-m class telescope employing adaptive optics to attain 0. 08. We also used the continuum far-IR images of the Herschel/Hi-GAL survey (Molinari et al. 2010) to accurately estimate the cluster luminosity. After describing the observational details in Sect. 2, we illustrate the results in Sect. 3, while the analysis and discussion of our findings is presented in Sect. 4. Finally, the conclusions are drawn in Sect. 5.

2. Observations and data reduction
2.1. Large Binocular Telescope
2.1.1. Near-IR images with LUCI

Near-IR images were taken in the night of December 12, 2009 with LUCI (Seifert et al. 2003) at the Large Binocular Telescope (Mount Graham, Arizona), through the standard broad-band filters H (c = 1.653 µm) and Ks (c = 2.163 µm), and the narrowband filters F eI I (c = 1.646 µm, including the [FeII] 1.64-µm line) and H2 (c = 2.122 µm, including the H2 v=1­0 S (1) line at 2.12 µm). We used the N3.75 camera with a pixel scale of 0. 12 and a field of view of 4 â4 . We took ten dithered images both at H and at Ks , each composed of 17 coadds of 2 s exposures (H ) and 15 coadds of 2 s exposures (Ks ). A set of 14 dithered images, each resulting from coadding five exposures of 20 s, were taken at H2 , whereas eight dithered images were taken at F eI I , each of them a coadd of nine exposures of 20 s. Total exposure times were then 300 s (Ks ), 340 s (H ), 1400 s (H2 ), and 1440 s (F eI I ), respectively. The dithering pattern consisted in alternatively imaging the target in the central part, the eastern part, and the western part of the detector field (i.e., with a throw of 1 in right ascension), with a random jitter of a few arcsec when the target was in the same area of the detector. These allowed us to obtain sky images from the on-source images themselves, by median filtering together 4 frames selected so that the target area did not overlap in the stack, which would have been a problem due to the extended emission feature in the field centre. All the images were flat-fielded, bad-pixel corrected, skysubtracted, registered, and combined by using standard IRAF1 routines. We note that, given the dithering strategy adopted, all final combined images attain their maximum signal-to-noise ratio in a central area 2 wide in right ascension. The full width at half maximum (FWHM) of the point-spread function (PSF) was between 0. 5 and 0. 6. We scaled the Ks and H2 final mosaics appropriately (by assuming a constant stellar continuum flux, which is correct only as a zero-order approximation) and then subtracted one from the other, to obtain a continuum-free map of pure H2 line emission. In practice, we solved the two equations that give the intensities measured in the H2 and Ks filters I
H2

= =

S S

H2 H2

c d + S c d + S

H2

t

H2



H2

(1) (2)

I

K

s

K

s

t Ks

K

s

H H c c with respect to S 2 d and S , where S 2 and S are the H2 line and continuum flux densities, H2 and Ks the widths of the H2 and Ks filters, tH2 and tKs the corresponding integration times, and H2 and Ks the transparencies of the two filters. We followed the same procedure with the H and F eI I final mosaics, but no [FeII] emission could be detected in the subtracted image, so we do not discuss the [FeII] data further. We performed PSF-fit photometry on the Ks and H final averaged (mosaiced) images by using the DAOPHOT routines in IRAF. In total, we found 1303 sources with detection at both Ks and H , 143 sources with detection at Ks and 60 sources with detection at H (see Table A.1). Unfortunately, the standard

IRAF is distributed by the National Optical Astronomy Observatory, which is operated by the Associated Universities for Research in Astronomy, Inc., under cooperative agreement with the National Science Foundation.

1


R. Cesaroni et al.: Star and jet multiplicity in the high-mass star forming region IRAS 05137+3919

3

Fig. 1. Maps of the continuum emission towards IRAS 05137+3919 at different wavelengths (given in the top right of each panel). The labels in the top, right panel indicate the three clumps/clusters that we refer to in the text. The images have been taken from our LUCI data (2.2 µm), the IRAC database (at 3.4 and 4.6 µm), the WISE archive (12 and 22 µm), the Herschel/Hi-GAL survey (70­250 µm), and Molinari et al. (2008) (850 µm). The contours in the bottom right panel are the map of the 3.6 cm continuum emission observed by Molinari et al. (2002), while the ellipse in the bottom right denotes the corresponding angular resolution. The circle in the bottom left of each panel represents the HPBW of the corresponding image.


4

R. Cesaroni et al.: Star and jet multiplicity in the high-mass star forming region IRAS 05137+3919

star observed for calibration during the night was saturated in all frames. Thus, we calibrated our photometry on 2MASS, by matching our sources to 2MASS PSC entries. For both bands, the colour range spanned by the common stars is too narrow to clearly show any colour effects between our instrumental magnitudes and 2MASS magnitudes, so we neglected colour terms. We obtained Ks (2MASS) - Ks (instrumental) = 23.81 ± 0.06 and H (2MASS) - H (instrumental) = 24.52 ± 0.05. The limiting magnitudes (at 3) are Ks 20 and H 20.5. A simple method of estimating the photometric completeness limits relies on histograms of magnitudes. Typically, the number of sources retrieved increases with increasing magnitudes up to a maximum value. Beyond this value, the source statistics is dominated by the decreasing efficiency in retrieving faint sources. Thus, the histogram maximum yields a rough estimate of the completeness limit. We conservatively assumed it to be 1 magnitude brighter than the histogram maximum. In fact, we found two nearby peaks at H , but the brighter one disappears when counting only sources in the 2 innermost part of the image. So the brighter peak is probably due to including sources from parts of the image with lower signal-to-noise ratio. The completeness limits we estimated are K s(compl) 17.75 and H (compl) 18.25. Finally, we calibrated the line emission fluxes on the final averaged (mosaiced) H2 image by repeating our photometry on the H2 image and matching it to the Ks calibrated photometry, correcting the stellar fluxes to the c of the narrow-band filter. Photometry of the H2 knots was then performed on the continuum-subtracted H2 images by means of task POLYPHOT in IRAF, enclosing each knot with polygons following the emission contour at 3 of the background counts as closely as possible. The identified knots and corresponding fluxes are given in Tables A.2 to A.4, where we have classified the knots on the basis of their association with Jet 1, Jet 2, and Jet 3 (defined in Sect. 3.3). For the H2 images the 3 sensitivity level is equal to 4 â 10-16 erg s-1 cm-1 arcsec-2 .
2.1.2. Near-IR images with FLAO and the PISCES camera

66 sources with detection both at Ks and H , 83 sources with detection at Ks , and 24 sources with detection at H (see Table A.5). We used aperture radii of 1 PSF-FWHM and annuli with inner and outer radii of 2­4 PSF-FWHM to derive the sky level, with the median as an estimator. We calibrated Ks and H by comparing the instrumental magnitudes to those obtained with LUCI (which in turn are calibrated on 2MASS) for the matching stars. Again, the colour range spanned by the data points is too narrow to show any clear colour effects, so we neglected colour terms. We obtained Ks (LUCI, 2MASS) - Ks (PISCES, instrumental) = 23.46 ± 0.10 and H (LUCI, 2MASS) - H (PISCES, instrumental) = 24.68 ± 0.11. We note that a few points depart by up to 0.4 mag from these relations. The most extreme differences are clearly due to unresolved objects in the LUCI images. Nevertheless, part of the scatter is probably due to PSF variations over the PISCES field (anisoplanatism), which is typical of AO-assisted images (e. g., Esslinger & Edmunds 1998). So, we can actually expect intrinsic errors of up to 0.1­0.2 mag in our PISCES photometry. The limiting magnitudes (at 3) are Ks 22.5 and H 21. We estimated the completeness limits as explained in Sect. 2.1.1. The distribution of Ks displays two nearby peaks (see Sect. 4.1), but the brightest one is likely to be intrinsic to the local stellar population. Thus, we derived Ks (completeness) 19.25 and H (completeness) 18.75. Finally, from the Ks and H2 images we obtained a continuum-free, pure H2 line emission map following the method described in Sect. 2.1.1.
2.2. Ground based mid-IR observations

The data were collected on October 9 and 12, 2013, using the PISCES Near Infrared Camera (McCarthy et al. 2001) installed at the focal plane of the First Light Adaptive Optics system (Esposito et al. 2010, 2011) of the Large Binocular Telescope. The detector has a pixel scale of 0. 02, with a field of view of 21 â21 . The observations were carried out through the H2 , H , and Ks filters. We used a star 10 west from our field center with R, I mag 12.0 as the reference for the AO loop, closed using 153 modes. The average Strehl Ratio on the centre of the field was 40% being measured on the H2 images. The electronic cross-talk between the quadrants in the PISCES Hawaii-I detector was corrected for each frame using Corquad, an IRAF procedure developed by Roelof de Jong2 . We obtained one image per band by registering and combining together, after sky subtraction: 119 exposures of 5 s at Ks (total integration time 595 s); 62 exposures of 25 s at H2 (total integration time 1550 s); and 3 exposures of 10 s at H (30 s). Due to the deteriorating weather conditions, only the latter small set of H frames is available. Each single frame was first drizzled, bad-pixel corrected, and flat fielded using standard IRAF routines. The PSF-FWHMs we measured are 0. 08 in all final images. We performed aperture photometry on the PISCES final images by using the DAOPHOT routines in IRAF. We found
2

See http://66.194.178.32/rfinn/pisces.html

Ground-based diffraction-limited mid-infrared images at 8.9, 9.9, 12.7, and 18.7 µm of IRAS 05137+3919 were taken on the night of November 9, 2006 with the mid-infrared camera CID (Salas et al. 2006) mounted on the 2.1 m telescope of the ´ ´ Observatorio Astronomico Nacional at San Pedro Martir, Baja California, Mexico. This camera is equipped with a Rockwell 128â128 square pixel Si:As BIB detector array that delivers an effective scale of 0. 55/pixel covering a fully-sampled area of 62 â62 . The images were taken with the standard chopnodding mode to remove the sky and telescope emission background. The standard stars Lyr, And, Aur, Her, and Aqu were observed before and after the programme sources at similar air masses for flux calibration, following Salas et al. (2006), in order to measure the PSF at each wavelength. These values (FWHM) ranged from 1.7 at the shortest wavelength, to 2. 1 at 18.7 µm. Individual images were obtained at ten nodding positions, 20 apart, while we chopped at 3 Hz with a throw of 22 . After all cycles, the total on-source integration time in each filter was 1440 sec. The astrometry of the images was determined by alignment with WISE and Spitzer/IRAC images. For this purpose, we smoothed the 8.9 and 9.8 µm images to 2 resolution and overlaid them by eye on the 4.5 µm IRAC image. The same procedure was adopted for the 12.7 and 18.7 µm images, which were compared to the 12 µm WISE image after smoothing them to 6. 5 resolution. Albeit not very accurate, this method results in an astrometrical error <1 , which will suffice for our purposes. A single mid-IR source was detected at all wavelengths. At < 13 µm, the source may be slightly resolved, with diameters >2 . The measured photometry on the CID calibrated images with a 4 aperture yields the following fluxes: 1.2 Jy at 8.9 µm, 1.5 Jy at 9.9 µm, 3.8 Jy at 12.7 µm, and 10.7 Jy at 18.7 µm. These values have 10% errors, dominated by uncertainties in the flux calibrations on standard stars as discussed by Salas et al. (2006).


R. Cesaroni et al.: Star and jet multiplicity in the high-mass star forming region IRAS 05137+3919

5

2.3. IRAM 30-m telescope

The observations were performed with the IRAM 30-m antenna on Pico Veleta (Spain) on July 13 and 14, 2003. The source was mapped with HERA, a multi-beam heterodyne dual polarization receiver, consisting of two arrays of 3â3 pixels with 24 spacing (Schuster et al. 2004). Areas of 4 â4 and 3 â3 centred around the position (J2000)=05h 17m 13s.8, (J2000)=39 22 20 were covered, respectively, in the 12 CO and C18 O J =21 rotational transitions. The maps were made in on-the-fly mode, scanning along the right ascension direction. The receiver was suitably rotated to allow for a sampling with 4 intervals in declination, perpendicular to the scanning direction, while the data were acquired along the scanning direction every 4 . This results in excellent sampling of the 12 halfpower beam width (HPBW) of the telescope. Position switch and frequency switch were used, respectively, for the 12 CO(2­ 1) and C18 O(2­1) line observations. The VESPA autocorrelator was chosen as a backend, with a spectral resolution of 0.1 km s-1 and 0.053 km s-1 , respectively, for the 12 CO and C18 O(2­1) lines. The pointing accuracy was regularly checked (typically every hour) on strong, pointlike continuum sources. The data presented in this paper are expressed in main beam brightness temperature, T MB , assuming a forward efficiency of 0.90 and a beam efficiency of 0.52 for 12 CO(2­1) and 0.55 for C18 O(2­1). The conversion to flux density, S , is given by the expression S (Jy) = 4.7 T MB (K). The data were reduced and analysed by means of the GILDAS software3 . The spectra were smoothed to 0.5 km s-1 before creating channel maps of the line emission.

3. Results
The observational findings of our study are illustrated in the following, where we concentrate on four main issues: the largescale structure, extending over 100 , or 4 pc, around the nominal position of IRAS 05137+3919 ­ (J2000)=05h 17m 13s.3, (J2000)=39 22 14 ; the small-scale structure, namely the region of <20 , or <0.8 pc, imaged with the PISCES camera; the jets and corresponding outflows (hereafter "jets/outflows") mapped both in the near-IR and at 3.4 mm; and the stellar clusters, spread over 2 .
3.1. The large-scale structure

For the sake of simplicity, in the following we refer to the northern clump and associated cluster (hereafter "clump/cluster") as C1, to the southern clump/cluster as C2, and to the south-eastern cluster as C3 (see the top, right panel of Fig. 1). C1, which is the target of our investigation, is by far the most luminous at all frequencies and is the only one where freefree emission has been detected by Molinari et al. (2002), as one can see from the contour map in Fig. 1. The richness of C1 is illustrated also in the left panel of Fig. 2, where a 3-colour image of the whole region obtained from the LUCI data is shown. Here, the presence of the three clusters C1, C2, and C3 stands out clear (white colour), as well as the existence of multiple H2 jets (red colour), which we will discuss in the following. Thanks to the Hi-GAL images, it is now possible to obtain an accurate estimate of the far-IR flux densities and hence of the luminosities of C1 and C2. We reconstructed the corresponding spectral energy distributions (SEDs) using data from the 2MASS (Skrutskie et al. 2006), Spitzer/IRAC (Werner et al. 2004; Fazio et al. 2004), IRAS, MSX (Price et al. 1999), WISE (Wright et al. 2010) archives, the Herschel/Hi-GAL survey, the SCUBA Legacy Surveys (Di Francesco et al. 2008), and the studies of Molinari et al. (2008) and Ishii et al. (1998). When possible, the flux densities were taken from the available source catalogues. In the other cases, for each clump we integrated the continuum flux inside a suitable polygon enshrouding all the emitting region, and then subtracted the background estimated through 1-D cuts across the clump centre. The flux densities computed in this way, as well as the solid angles encompassed by the polygons used for the integration are given in Table A.6, while the corresponding SEDs are shown in Fig. 3. The bolometric luminosity was estimated by integrating the SED over the whole wavelength range, after linear interpolation in the Log­LogS plot. For the fiducial distance of 8.3 kpc we obtain 2.4 â 104 L for C1 and 6.7 â 103 L for C2. These numbers agree within a factor 2 with the estimates obtained from the cluster simulations of Faustini et al. (2009) and Kumar et al. (2006), once the different distance assumed in their calculations (11.5 kpc) is taken into account.
3.2. The small-scale structure

As already noted by various authors, two molecular clumps are detected in the IRAS 05137+3919 region, offset by 30 along the N­S direction. In this study we focus on the northern one, but it is worth mentioning that the two clumps are likely physically close in space and not only in projection onto the plane of the sky, because their LSR velocities differ by only 1 km s-1 (see Brand et al. 2001). Each of them hosts a stellar cluster, as pointed out by Faustini et al. (2009). In Fig. 1 we show a number of maps of the continuum emission of IRAS 05137+3919, ranging from the near-IR to the submm. We have chosen the images with the best angular resolution available in each wavelength regime. Besides confirming the previous findings, the IR images clearly indicate the existence of a third stellar cluster, located 15 to the east of the southern clump and considerably fainter than the other two clusters, which are dominating the emission at all wavelengths.
The GILDAS software has been developed at IRAM and Observatoire de Grenoble ­ see http://www.iram.fr/IRAMFR/GILDAS
3

As explained in Sect. 2, with the LBT/PISCES imaging we have attained a resolution of 80 mas or 664 au over 20 . This allows us to investigate in detail the innermost region of C1, which is the target of the present study. Our findings are best illustrated by the 3-colour image in the right panel of Fig. 2, where several features are worth of mention. The "white" star to the west of the image, is very likely a foreground object, not physically related to the cluster of interest for us. The two "yellow" stars at the centre are well separated, with the one to the NE (hereafter A, after Varricatt et al. 2010) appearing slightly "redder" and hence more extincted than the other (hereafter B). While the presence of this pair was already known, in our image A splits into two sources, one, brighter, to the SW (which we call A1) and another, fainter, to the NE (A2). This is more evident in Fig. 4, where we compare the LUCI to the PISCES images. Clearly, the dramatic improvement in angular resolution obtained with AO is crucial to resolve A1 from A2, whose apparent separation is 0. 18 (i.e. 1500 au). All three sources, A1+A2+B, are likely responsible for the ´ mid-IR emission measured by us with the San Pedro Martir telescope, as demonstrated by Fig. 5, where we overlay the LUCI Ks image on the four images at wavelengths ranging from 8.9 to 18.7 µm. Despite the limited astrometrical accuracy (1 ), it is


6

R. Cesaroni et al.: Star and jet multiplicity in the high-mass star forming region IRAS 05137+3919

Fig. 2. Composite colour images obtained by combining the emission in the H2 (red), Ks (green), and H (blue) filters from the LUCI (left panel) and PISCES (right panel) data. The dashed box in the left panel indicates the region shown in the right panel. The LUCI and PISCES images cover regions, respectively, of about 0. 8â1. 1 and 0. 3â0. 4.

Fig. 3. Spectral energy distributions of clumps/clusters C1 (left panel) and C2 (right panel). The colour coding of the symbols is explained in the left panel. The triangles indicate upper limits.

clear that the bulk of the mid-IR emission arise from the same region where the near-IR emission peaks. In turn, this implies that the three sources are responsible for most of the luminosity estimated for C1 (2.4 â 104 L ), because their mid-IR emission at 12.7 and 18.7 µm corresponds to 80% of the flux densities measured within much greater HPBWs with MSX and WISE (see Table A.6).

An interesting feature seen in Fig. 2 is the "green" arc of continuum emission extending to the SW from B. This almost coincides with three H2 knots, which can be recognised from their red colour in the image. Whether the arc and the knots are physically related is difficult to establish, but it is possible that both are manifestations of the same phenomenon, perhaps a bow shock due to Jet 1. In any case, the knots are in all likelihood due to the interaction between the jet and the dense gas associated


R. Cesaroni et al.: Star and jet multiplicity in the high-mass star forming region IRAS 05137+3919

7

Fig. 4. Images of the central region of clump/cluster C1, where a pair of bright stars (A and B) had been detected in previous studies. The left and right panels refer, respectively, to Ks and H band images, while the upper and lower panels show, respectively, images obtained with the LUCI and PISCES+AO cameras. The values of the contour levels are indicated by marks in the corresponding colour scales. Note how employing AO allows us to resolve star A into the two sources A1 and A2.

with C1. As explained in Sect. 1, Jet 1 was first observed by Varricat et al. (2010) and is confirmed by our images (see below). Finally, at the top and bottom of the right panel of Fig. 2, one sees two other "red" regions of H2 line emission, shaped like bow shocks. These correspond to (part of) Jet 2, also identified by Varricat et al. (2010).
3.3. The bipolar jets/outflows

Evidence for the existence of multiple jets/outflows in the IRAS 05137+3919 region has been provided by various studies, as described in Sect. 1. Our high-resolution images shed new light on this issue, thanks to their superior sensitivity and angular resolution. Moreover, the velocity information conveyed by the 12 CO maps makes it possible to discriminate between the blueand red-shifted lobes. In Fig. 6, maps of the outflow lobes are shown, obtained by integrating the 12 CO(2­1) emission over the line wings, in various velocity i