A multi-wavelength study of the star-forming core ahead of HH 80N
Josep-Maria Masqué, Mayra Osorio, Josep-Miquel Girart, Guillem Anglada, Guido Garay, Robert Estalella, Nuria Calvet, Maria-Teresa Beltrán
aa r X i v : . [ a s t r o - ph . S R ] J un Printed November 17, 2018
A multi-wavelength study of the star-forming core ahead of HH80N
Josep M. Masqué , Mayra Osorio , Josep M. Girart , Guillem Anglada , Guido Garay ,Robert Estalella , Nuria Calvet andMaria T. Beltrán ABSTRACT
We present observations of continuum emission in the MIR to mm wavelengthrange, complemented with ammonia observations, of the dense core ahead of theradio Herbig Haro object HH 80N, found in the GGD 27 region. The continuumemission in all the observed bands peaks at the same position, consistent withthe presence of an embedded object, HH 80N-IRS1, within the core. The distri-bution of the VLA ammonia emission is well correlated with that of the dust,suggesting that photochemical effects caused by the nearby Herbig Haro objectdo not play an important role in shaping this particular molecular emission. Inorder to unveil the nature of HH 80N-IRS1 we analyzed the continuum data ofthis source, using self-consistent models of protostellar collapse. We find that ayoung protostar surrounded by a slowly rotating collapsing envelope of radius ∼ M ⊙ plus a circumstellar disk of radius ∼
300 AU and 0.6 M ⊙ Departament d’Astronomia i Meteorologia, Universitat de Barcelona, Martí i Franquès 1, 08028Barcelona, Catalunya, Spain Instituto de Astrofísica de Andalucía, CSIC, Glorieta de la Astronomía S/N, E-18008 Granada, Spain Institut de Ciències de l’Espai, (CSIC-IEEC), Campus UAB, Facultat de Ciències, Torre C5 - parell 2,08193 Bellaterra, Catalunya, Spain Departamento de Astronomía, Universidad de Chile, Camino el Observatorio 1515, Las Condes, Santiago,Chile Department of Astronomy, University of Michigan, 830 Dennison Building, 500 Church Street, AnnArbor, MI 48109, USA INAF - Osservatorio Astrofisico di Arcetri, Largo E. Fermi 5, 50125 Firenze, Italy µ m, 1.2 mm and 3.5 mm of HH 80N-IRS1. Besides, the APEX and PdBIcontinuum maps at 350 µ m and 3.5 mm, respectively, reveal additional clumps inthe continuum emission. Given the modeling results and the observed morphol-ogy of the emission, we propose a scenario consisting of a central embedded Class0 object, HH 80N-IRS1, with the rest of material of the HH 80N core possiblyundergoing fragmentation that may lead to the formation of several protostars.
1. Introduction
The region of GGD 27 (Gyulbudaghian, Glushkov, & Denisyuk 1978) in Sagittarius ata distance of 1.7 kpc (Rodríguez et al. 1980), is an active star forming region with manyproperties still unveiled. The most well-known observational signature of this region, theHH 80/81/80N jet, is one of the largest collimated jet systems known so far, expanding overa total length of ∼ − ∼ M ⊙ ), and compact( r . AU) disk. Copious UV radiation, generated in the Herbig Haro (HH) shocksand the jet, is able to induce the formation of a Photodissociation Region (PDR) along thebipolar flow (Molinari et al. 2001).Ahead of the radio source HH 80N, the obscured northern head of the jet, there is adense core of ∼ and HCO + emission led Girart etal (1998) to suggest an unusual chemistry for this core. From the ammonia emission of thecore a mass of roughly 20 M ⊙ was estimated by Girart et al. (1994).Claims for evidence of association of molecular clumps with HH objects have beengrowing over the last decades (Rudolph & Welch 1988; Davis, Dent & Bell Burnell 1990;Torrelles et al. 1992, 1993; Girart et al. 1994, 1998; 2002, 2005; Viti, Girart & Hatchell 2006;Whyatt et al. 2010). These clumps do not show evidence for shock heating as indicatedby their typically observed narrow linewidths (< 1 km s − ), low temperatures (< 20 K)and radial velocities close to the ambient cloud velocity (e.g. Whyatt et al. 2010). Allthese properties rule out the possibility of a dynamical perturbation. A possible explanationfor the association of molecular clumps with HH objects, proposed by Wolfire & Königl 3 –(1992), is that the molecular clump is exposed to a strong UV radiation generated in theHH objects. This radiation is capable of evaporating icy mantles on dust grains, triggeringa non-equilibrium chemistry that leads to an increase in abundance of specific molecularspecies. This scenario was extensively modeled by Taylor & Williams (1996) and Viti &Williams (1999).In the frame of the photo-illuminated scenario, Girart et al. (1998) suggest that theobserved chemistry of the HH 80N core is compatible with the radiative shock-inducedchemistry models, being HH 80N the most plausible source of UV irradiation. A morerecent study, using several molecular species, appears to be in agreement with this chemicalscenario, even though the photochemical effects are observed tentatively only in the side ofthe core facing HH 80N (Masqué et al. 2009). However, the mass and size of the HH 80Ncore make this clump quite different from the other molecular clumps found ahead of HHobjects. The latter are smaller and less massive (size ≤ M ≃ M ⊙ ) than the HH80N core and do not show any signpost of star formation. Indeed, they are believed to betransient structures or small density fluctuations within the molecular clouds (Viti et al.2003; Morata et al. 2003, 2005).Interestingly, Girart et al. (2001) found evidence of a bipolar CO outflow centered nearthe peak of the HH 80N molecular core, suggesting the presence of an embedded protostar.In addition, Girart et al. (2001) and Masqué et al. (2009) interpreted the distribution andkinematics of CS and other molecular species observed in the outer parts of the HH 80Ncore as suggestive of an infalling ring-like molecular structure (probably caused by strongmolecular depletion in the inner parts) with a radius of ∼ ∼ . km s − ), significantly larger than those expected in the standard protostellar collapse(e.g. ambipolar diffusion models predict gas inward motions of a fraction of the isothermalsound speed, ∼ − , at scales similar those of the HH 80N core; Basu & Moscouvias1994), led these authors to suggest a peculiar dynamical evolution of this core. Given theparticular environment of the HH 80N core, we question whether the HH 80/81/80N outflowhas triggered or at least sped up in some way the collapse of the core.In order to unveil the nature of the HH 80N core, we carried out an extensive set ofcontinuum and ammonia observations over the last years using several instruments, amongthem VLT and APEX. Using radiative transfer models of collapsing protostellar envelopes,we compare the synthetic spectral energy distribution (SED) and spatial intensity profiles ofthe dust emission with the data. From this comparison we derive the mass and luminosityof the collapsing envelope, mass infall rate, and mass of the embedded object under thehypothesis that a young stellar object (YSO) is forming inside of the HH 80N core. Weexplore several density profiles for the initial configuration of the collapsing envelope, and 4 –discuss whether the inclusion of an accretion disk is needed to simultaneously explain theobserved emission at different wavelengths. Our results point to a new picture for the HH80N core, which consists in a ’classic’ Class 0 object (hereafter, HH 80N-IRS1) surroundedby a reservoir of material where other clumps potentially forming stars are found.The paper is organized as follows. In Section 2, we describe the observations. InSection 3 we present the resulting maps. In Section 4 we model the observed dust emissionusing different approaches for the protostellar collapse. The discussion of the results of themodeling is done in Section 5. Finally, our conclusions are summarized in Section 6.
2. Observations2.1. VLA
We observed the (J,K) = (1,1) and (2,2) ammonia inversion transitions (at 23.6944955 GHzand 23.7226336 GHz, respectively) using the Very Large Array (VLA) of the NRAO inthe D configuration on January 31, 2007. The 4IF spectral correlator mode was used,which allows to observe both transitions in two polarizations simultaneously. The corre-lator was set to observe a bandwidth of 1.56 MHz with 63 spectral channels of 24.4 kHz(which gives 0.115 km s − of velocity resolution at 1.3 cm) plus a continuum channel thatcorresponds to 75% of the bandwidth. The phase center of the observations was set to α ( J h m . s and δ ( J − ◦ ′ . ′′ . The flux calibrator was 3C286 witha flux of 2.59 Jy, the phase calibrator was 1832-105 with a bootstrapped flux of . ± . Jy and the bandpass calibrator was 0319+415. Maps of the ammonia (1,1) transition wereobtained with natural weighing and using a Gaussian taper of 35 k λ , which gives a beam sizeof . ′′ × . ′′ (P.A. = 20.0 ◦ ) and an rms noise level of × − Jy beam − per channel. Herewe only present the map of the velocity-integrated flux density of the (1,1) transition, whichpresents extended emission with a good signal to noise ratio (see § 3.2), for a comparisonwith the continuum data. A more complete analysis including the ammonia (2,2) transition,that is sensitive to warmer gas, will be presented in a future paper. The National Radio Astronomy Observatory is a facility of the National Science Foundation operatedunder cooperative agreement by Associated Universities, Inc.
The 1.2 mm continuum observations were carried out with the 117-channel bolometerarray (MAMBO 2) installed at the 30m IRAM telescope (Pico Veleta, Spain) on January27 and 28 and March 3, 2008. The data were taken under low sky-noise conditions withthe zenith opacity at 250 GHz ranging from 0.14 to 0.28. The on-the-fly technique wasused with a scanning speed of ′′ s − . Chopping was performed with throws of ′′ and ′′ .The angular resolution of the map is . ′′ and the achieved rms noise level at this angularresolution is 1.5 mJy beam − . Data reduction was carried out with the MOPSIC software. The Plateau de Bure Interferometer (PdBI) observations were carried out on April3 and 10, 2010 in the C configuration. Two tracks were performed under good weatherconditions. We made a 6-point mosaic with the phase tracking center of the observations setat α ( J h m . s and δ ( J − ◦ ′ . ′′ , coincident with the Spitzer 8 µ mpeak position (see § 3.1). The receivers were tuned to the rest frequency of the N D (1 , –1 , )line. The correlator was configured in four widex units (two units for each polarization) thatgives a total continuum bandwidth of 8 GHz, plus two narrow windows of 20 kHz coveringthe N D (1 , –1 , ) (85.926263 GHz) and HN C (1–0) (87.090859 GHz) transitions, with avelocity resolution of ∼ . km s − . The phase calibrator was QSO 1911-201, which hasa flux density of 1.09 Jy at the observed frequency, and the flux and bandpass calibratorwas MWC349. The data were calibrated using CLIC following the baseline-based mode,since antenna-based solutions were not optimal for the shortest baselines. The continuummap was obtained using MAPPING with natural weighting, that gives a synthesized beam(HPBW) of . ′′ × . ′′ (P.A. = ◦ ) at 86 GHz. The spectral line data will be presented ina future paper. In the nights of May 2 and June 12, 2009, Q-band (20 µ m) imaging of the HH 80N regionwas carried out in service mode using the VISIR instrument installed at the Cassegrain focusof the UT3 telescope (Melipal) of the Very Large Telescope (VLT). The observations covereda field of view of . ′′ × . ′′ at a pixel scale of . ′′ /pixel.The sky conditions were good: for the night of May 2 the optical seeing ranged from . ′′ to . ′′ and the air-mass was ∼ . ′′ to . ′′ and the air-mass was always less than 1.05. Under theseconditions, the extrapolated seeing at Q band according to the Roddier formula ( ∝ λ . )is < . ′′ . Taking into account that the VLT diffraction limit at the same band is . ′′ , thisimplies that the angular resolution of our Q band data is basically dominated by diffraction.During the first observing night our target was observed for one hour through theQ1, Q2, and Q3 filters (covering . ± . µ m, . ± . µ m, and . ± . µ m,respectively). However, due to the low luminosity of the target, it was not detected in anyof these filters. In the second night, our target was observed solely in the Q2 filter for onehour and was detected with a signal-to-noise ratio of 17. In order to remove the atmosphericand telescope background the standard chopping and nodding technique in perpendiculardirections was carried out with chop-throws of ′′ . The standard star for photometriccalibration, HD178345 (3.52, 3.15 and 2.88 Jy through Q1, Q2 and Q3 filters, respectively),was observed immediately after our target. A preliminary reduction of the data was carriedout using the standard ESO reduction software including the graphical user interface to thepipeline, GASGANO. The final image resulting of shifting and combining the chopping andnodding cycles was obtained with the IRAF package.The photometric calibration was performed using the ’visir_img_phot’ recipe with thecombined image of the cataloged standard star as an input. We obtained conversion factorsof 11923.3 (Q1), 11235.0 (Q2, first night), 13379.4 (Q2, second night) and 2446.2 (Q3)between the number of detector counts per second and the source flux in Jy. Q2-band fluxdensity of the target was obtained with standard aperture photometry using the PHOT taskof IRAF with a circular aperture of radius . ′′ although we inspected values between . ′′ and . ′′ . For the background subtraction, the sky contribution was fitted to an annulus situatedbetween radii of . ′′ and . ′′ from the center of the aperture. Applying the conversion factorderived above, we obtain a flux density value of 0.175 mJy at 18.7 µ m. The flux uncertaintywas derived exploring the variation of the flux density value when measured using differentapertures. The sub-mm data were obtained with the SABOCA camera, a 39-pixel bolometer arraylocated on the Atacama Pathfinder EXperiment (APEX) telescope in the Chilean Andes.Each SABOCA pixel consists in a composite bolometer with superconducting thermistor onsilicon-nitride membranes. The pixels are arranged in a hexagonal layout consisting of acentral channel and 3 concentric hexagons. The array is installed at the Cassegrain focus,where it has an effective field of view of ′′ . SABOCA operates at 850 GHz (350 µ m) which 7 –gives a beamsize (FWHM) of . ′′ . The observations were carried out on October 7, 2009 inraster spiral mode with 4 scans, each providing a fully sampled area of ∼ ′′ × ′′ . Twoskydips, taken in-between and after on-source observations, yielded values of ∼ . ′′ Gaussian that gives an angularresolution of . ′′ and an rms noise of 0.09 Jy beam − for the final map. The IRAS Point Source Catalog (PSC) reports a weak source, IRAS 18163-2042, whoseposition uncertainty includes the HH 80N core. According to the catalog, this source wasdetected only at 60 µ m with a flux density of ∼ ≤ µ m, and ≤
250 Jy at 100 µ m). However, thesevalues are not reliable because of the presence of strong side lobes generated by the nearbyluminous source IRAS 18162-2048. These side lobes create a complex background aroundIRAS 18163-2042, which may cause an underestimate of the flux values reported in the IRASPSC. This motivated us to SCANPI reprocess the four IRAS bands covering the HH 80Nregion. SCANPI, a utility provided by the Infrared Processing and Analysis Center (IPAC),performs 1-dimensional scan averaging of the IRAS raw survey data. As it combines all thescans passed over a specific position, its outcome has a higher sensitivity, ideal to obtainfluxes of confused or faint sources. In particular, among all the possible input processingparameters of SCANPI, we gave special attention to the ‘local background fitting range’. Bydefault, an interval of radius ′ from the scan center is used to fit the background. From thisinterval, a central range depending on the band is excluded in the fitting in order to preventcontamination from the target: ′ (12 µ m and 25 µ m), ′ (60 µ m) and ′ (100 µ m). Using thisdefault SCANPI yields values similar to the fluxes of the IRAS PSC for all the 1-D scans thatpass thorough the HH 80N region. However, IRAS 18162 − ′ south-east fromIRAS 18163 − ′ for the ‘source exclusion range’in the background fitting for all the bands, we obtained significantly higher flux values thanthose derived above. Other values for the ‘source exclusion range’ yielded similar resultsprovided that IRAS 18162 − µ m IRAC bands that are part of a large sample of high-mass star forming regions(PID: 3528). The data at these wavelengths complement very well our VLT observations.Following the same procedure as for the Q band data, we estimated the flux density at4.5 and 8 µ m with the aperture photometry technique using PHOT of IRAF. We found thatsignificantly larger apertures must be used for the Spitzer data compared to the aperturesused for the VLT data: ′′ and ′′ for and 4.5 and 8 µ m images, obtaining flux values of0.026 and 0.041 Jy for 4.5 and 8 µ m, respectively. As for the VLT data, the flux uncertaintywas derived by monitoring the variation of the flux value when measured using differentapertures.Finally, we made use of the recently available data from the Infrared AstronomicalSatellite Akari. Akari is equipped with two instruments, IRC, covering several bands at nearand mid-infrared wavelengths, and FIS, covering several bands at far-infrared wavelengths.From a total of six observable bands, in this paper we present the data of 18, 140 and 160 µ mbands. The rest of bands have bad quality data or null detection for our source.
3. Results3.1. Continuum emission
Figure 1 presents the mid-IR images of the HH 80N region. The Spitzer images at 4.5and 8 µ m show a bright compact source dominant in the two wavelengths (HH 80N-IRS1).Two other compact sources, located ∼ ′′ South-East (HH 80N-IRS2) and ∼ ′′ North-West (HH 80N-IRS3) of HH 80N-IRS1, may belong to the region. The positions of thesecompact sources are given in Table 1. HH 80N-IRS1 is also detected in the VLT imageat 18.7 µ m, but without any extended structure, probably due to the short exposure timeof our observations. Taking into account that the source is not resolved, and given theangular resolution of VLT at Q-band ( ∼ . ′′ ), the warm part of the envelope remains within ∼
425 AU (for 1.7 kpc of distance) from the core center. 9 –HH 80N-IRS1 is located at the center of the HH 80N core that is seen in absorptionagainst the emission of the Galactic background in the 8 µ m image. This observationalpicture resembles that of InfraRed Dark Clouds (IRDC), being the size of the silhouetteof the HH 80N core ( ∼ µ m emission is elongated with an angular size of ′′ × ′′ (FWHM) and P.A. ∼ ◦ , and peaks at the position of HH 80N-IRS1. Apart from HH 80N-IRS1, the 350 µ m emission traces additional material towards the South-East. The 1.2 mmemission peaks at the same position and traces fairly well the silhouette of the absorptionfeature of the 8 µ m image including the north-western tail expanding up to ′′ from thecentral peak. The detection of this tail in emission at 1.2 mm and in absorption at 8 µ mexcludes the possibility of being an artifact of the 1.2 mm map. In the PdBI 3.5 mm mapthe dust emission splits into two main sources, one clearly associated with HH 80N-IRS1,plus another southeastern component (hereafter Southeastern Condensation). Table 2 givesthe results of Gaussian fits of the 3.5 mm emission for HH 80N-IRS1 and the SoutheasternCondensation. In addition, the 3.5 mm map shows two marginally detected sources located ∼ ′′ northwest and ∼ ′′ southeast of HH 80N-IRS1.Table 3 gives a summary of the flux density measurements towards HH 80N-IRS1. Sinceat mm and submm wavelengths it is difficult to discriminate which fraction of the emissionof the HH 80N core corresponds to HH 80N-IRS1, in the table we report a range of possiblevalues for the flux densities at these wavelengths. The upper limit of the range at 1.2 mmand 350 µ m is an estimate of the HH 80N-IRS1 flux density avoiding the contamination fromthe Southeastern Condensation. To do this, we integrated the flux density of the western halfof the HH 80N core and multiplied the resulting value by 2. The lower limit is the intensitypeak that would coincide with the flux density of HH 80N-IRS1 if it were an unresolvedsource (i.e. the lowest possible contribution). For the 3.5 mm measurement, we take inaccount the missing short spacings of the PdBI. We made a crude analysis simulating thefiltering effects of the u - v coverage of our PdBI observations. Using the UVMODEL taskof MIRIAD package, we tested these filtering effects on several synthetic maps of artificiallygenerated ellipses that mimic the HH 80N core appearance. We find that a maximum of a50% of the total flux is missed. Thus, in the range of flux densities given for the 3.5 mmemission, the lower limit corresponds to the flux density measured in the map and the upperlimit corresponds to this value corrected by a factor of 2. The data obtained with lowangular resolution (i.e. all the IRAS data and the Akari 140 and 160 µ m bands) are likely 10 –contaminated by background sources. Therefore, we considered these fluxes as upper limits.Finally, note that the complex background of the 8 µ m image (see Fig. 1) makes the estimateof the flux at this wavelength somewhat uncertain. emission Figure 2 (bottom panel) presents the NH (1,1) emission superimposed to the Spitzer8 µ m image. The ammonia is very well correlated with the 8 µ m absorption feature. Itis also well correlated with the 1.2 mm emission. This is not the case for other moleculartracers such as CS or SO presented in previous works (Girart et al. 2001; Masqué et al.2009), whose emission is significantly more extended ( ∼ ′′ × ′′ ). These studies also showthat, in general, the molecular tracers do not peak all at the same position, probably becausethese molecules are depleted in the densest and inner part of the HH 80N core, that is welltraced by the dust continuum emission.The clear correlation between the NH emission and dust emission indicates that NH traces fairy well the material of the HH 80N core. Girart et al. (2001) detected star formingsignatures in the core such as a bipolar outflow traced by CO. In addition, the dust continuumemission shows a compact source, HH 80N-IRS1, in all the observed wavelengths, suggestingthe presence of an embedded YSO. All these results suggest that the HH 80N core is currentlyundergoing active star formation, and that the observed distribution of NH emission arisesas a consequence of the high gas densities likely reached in the core, similarly to other starforming cores, and not as a consequence of photochemical effects. Ammonia emission arisingas a consequence of a dynamical perturbation is excluded by the narrow NH (1,1) linewidth( ∼ − ) observed. Nevertheless, we note that the strongest NH emission is found inthe Southeastern part of the core, close to HH 80N, coinciding with the emission detectedin the lower sensitivity ammonia observations of Girart et al. (1994). This could be due toa local increment of abundance in this part of the core as found in some species (Masqué etal. 2009); indeed, NH is one of the species predicted to be enhanced by the HH radiation(Viti et al. 2003). Understanding these local departures of the ammonia emission from theglobal distribution of gas and dust in the HH 80N core is an issue that will require furtherinvestigation. 11 –
4. Modeling
In the following we analyze this region assuming that an YSO is forming inside the HH80N core. To do that, we calculate the dust emission arising from an envelope of dust andgas that is collapsing onto a central star. We consider three possible density profiles for theenvelope. We first investigate the collapse of a Singular Logatropic Sphere (SLS, McLaughlin& Pudritz 1996; Osorio et al. 1999, 2009). The SLS has a logarithmic relationship betweenpressure and density, introduced by Lizano & Shu (1989) to empirically take into account theobserved turbulent motions in molecular clouds. In the SLS collapse solution an expansionwave moves outward into the static core and sets the gas into motion toward the central star.Outside the radius of the expansion wave the SLS envelope is static, with a dependence ofthe density with radius as ρ ∝ r − . Inside the radius of the expansion wave the gas fallsonto the central star with a nearly free-fall behavior ( v ∝ r − / , ρ ∝ r − / ) at small radii.As a second approach, we adopt the collapse solution of the Singular Isothermal Sphere(SIS, Shu 1977). In this case the collapse occurs in a similar fashion than in the logatropic casebut the radial dependence of the density in the static region goes as ρ ∝ r − . Nevertheless,there are important differences in the evolution of both types of collapse as both the speedof the expansion wave and the mass infall rate are constant in the SIS collapse while theyincrease with time in the SLS collapse.As a third approach we use the solution for the collapse of a slowly rotating core de-scribed in Terebey, Shu and Cassen (1984), hereafter the TSC collapse (see also Cassen &Mossman 1981; Kenyon et al. 1993). In this model, the initial equilibrium state correspondsto the uniformly rotating analogue of the SIS. To first order, the collapse proceeds similarlyto the isothermal case beginning at the center of the core and propagating outward at thesound speed as an expansion wave. Material outside the radius of the expansion wave re-mains in hydrostatic equilibrium (with ρ ∝ r − ) while inside this radius the infall velocityand density approach those of free fall. However, the angular momentum of the infalling gasbecomes important in the vicinity of the centrifugal radius, where motions become signifi-cantly non radial and material then falls onto a circumstellar disk rather than radially ontothe central object. The centrifugal radius is given by R c = r Ω / ( GM ∗ ) , where Ω is theangular velocity at a distant reference radius r .The HH 80N core has a moderate bolometric luminosity, but relatively strong mmand submm emission. By integrating the area below the observed SED, constructed withthe flux densities of Table 3, we can derive a possible range of luminosities for HH 80N-IRS1. Considering the lower limits of the mm and submm points and excluding the rest ofcontinuum data in this calculation, we obtain 10 L ⊙ as the luminosity lower limit. Similarly,taking the upper limits of the mm and submm range and including the 60 µ m IRAS point, 12 –which is the most restrictive upper limit in the IR part of the SED, we obtain an upper limitof 110 L ⊙ for the luminosity. We are assuming that the total luminosity, which is responsiblefor internal heating of the core, is the sum of the stellar luminosity and the infall luminositycaused by the infalling gas onto the protostar. The stellar luminosity can be related to themass of the central star using the Schaller et al. (1992) evolutionary tracks. The upper limitof the luminosity range deduced above (110 L ⊙ ) restricts the mass of the central embeddedobject to ≤ M ⊙ , according to the tables of Schaller et al. (1992); this mass upper limitcorresponds to the hypothetical case that all the luminosity of the source was due entirelyto the stellar luminosity.For the SLS and SIS envelopes, the dust temperature is self-consistently calculatedfrom the total luminosity using the dust opacity and the procedures described in Osorio etal. (1999, 2009). These authors calculate the dust opacity at short wavelengths ( λ < µ m) assuming that the dust in the envelope is a mixture of graphite, silicates, and waterice, with abundances taken from D’Alessio (1996), and assuming a power law of the form κ λ ∝ λ − β , with ≤ β ≤ , for λ ≥ µ m.The temperature in the TSC case is also self-consistently calculated from the totalluminosity, following the procedures described in Calvet et al. (1994) and Osorio et al.(2003). The latter authors obtain the dust opacity over the whole wavelength range assuminga mixture of graphite, silicates and water ice, whose parameters (grain size and abundance)are obtained by fitting the well sampled SED of the prototypical class I object L1551 IRS5.We computed the SEDs of the models and compared them with the observed values ofthe flux density of HH 80N-IRS1 (Table 3). Additionally, we produced synthetic maps ofthe model emission at 3.5 mm, 1.2 mm, and 350 µ m bands. In order to fully simulate theobservations, the synthetic maps at 1.2 mm and 350 µ m were convolved with Gaussians withFHWM of . ′′ and . ′′ , respectively. For the synthetic map at 3.5 mm, the effect of themissing short spacings of the interferometric observations was taken in account. We used theUVMODEL task of MIRIAD to compute the visibility tables of the 3.5 mm models with thesame u - v plane coverage as our PdBI observations. Then, from the model visibility tables,we obtained synthetic maps following the standard data reduction routines of the MIRIADpackage.Finally, to check the goodness of our results, we compared selected models with thedata by means of spatial intensity profiles obtained using the task CGSLICE of the MIRIADpackage, for both the synthetic and observed maps at 1.2 mm, 3.5 mm, and 350 µ m. Wepresent two cuts of the observed intensities with P.A. ≃ ◦ and ≃ ◦ (mainly along themajor and minor axes of the HH 80N core seen at 1.2 mm and 350 µ m). An intensityprofile obtained with a cut with a selected PA, instead of averaging the emission over an 13 –annulus, prevents to include contamination of the Southeastern Condensation. In practice,the synthetic maps of the models in the three approaches have radial symmetry since, in theTSC case, the angular scale of the flattening of the envelope is small enough that it onlyaffects the central pixel of the map. Therefore, for the synthetic maps, we obtained a cutalong the diameter of the modeled source.We explore the SLS and SIS cases separately by running a grid of models taking themass infall rate ( ˙ M i ), mass of the central embedded object ( M ∗ ) and the external radius ofthe core ( R ext ) as free parameters. Because of the complex morphology of the source (seeFig. 2) we do not adopt a fixed value of R ext ; rather, its value is constrained in the fit. Theopacity index ( β ) was derived to get a trade off to reproduce simultaneously the emission at1 mm and 350 µ m. This yields β ∼ . for the SLS case and β ∼ . for the SIS case. Forthe stellar radius ( R ∗ ) we chose a standard value of 5 R ⊙ (Schaller et al. 1992). Given thespace of parameters ( R ext , ˙ M i ), we tested values of M ∗ from 0.5 to 1 M ⊙ . Because a fractionof the luminosity is due to infall, higher values of M ∗ would yield luminosities that exceedthe upper limit of 110 L ⊙ derived above, assuming that ˙ M i > − M ⊙ yr − .The best fit model can be determined by calculating the χ -statistics obtained from theresidual map resulting by subtracting the synthetic from the observed maps. This processwas performed for all the sub/mm bands. To find the best fit to the images, we do notinclude contamination of the Southeastern Condensation in the χ analysis. For the 350 µ mband, the χ function was calculated over a region comprising only the western part of theHH 80N core, roughly a semicircle with a radius of ∼ ′′ . In the 1.2 mm map, HH 80N-IRS1cannot be separated from the Southeastern Condensation due to the lower angular resolutionof the IRAM 30m observations with respect to the APEX observations. For this case we fitthe intensity profile obtained along the minor axis of the emission (i.e. towards NE). For the3.5 mm map, as HH 80N-IRS1 appears detached from the Southeastern Condensation in themap, a box of ∼ ′′ × ′′ enclosing the entire source was used. Additionally, we calculatedthe χ function for the SED by comparing the fluxes of Table 3 with the predicted SED ofthe model.Because the TSC models predict the development of a flattened rotating structure atthe innermost part of the envelope, we carried out the modeling assuming a TSC envelopefalling onto a disk surrounding the central object. We browsed an appropriate disk modelfrom the online catalog of models of irradiated accretion disks around pre-main sequencestars (D’Alessio et al. 2005). For consistency, the orientation of the disk is chosen to coincidewith the rotation axis of the TSC envelope and the disk radius is fixed to the value of thecentrifugal radius R c . To obtain the SED of the total composite model, we added the fluxesof the disk and envelope at each wavelength accounting for the extinction of the envelope, 14 –which is important at short wavelengths. Since the disk is unresolved even in the imageswith the highest angular resolution, to obtain the synthetic maps we added the disk to theenvelope as a point source at the central pixel.Due to computational limitations, the fitting process of the TSC envelope plus circum-stellar disk could not be automated. Our strategy, then, was to perform a case-by-caseexploration varying the external radius of the core ( R ext ), the radius of the expansion wave( R ew ), and the reference density ( ρ ) (this latter parameter is the density the envelope wouldhave at a radius of 1 AU for the limit R c = 0 , and it is related to the mass infall rate andthe central mass through equation 3 of Kenyon et al. 1993), until we found a model thatexplains satisfactorily the SED and the intensity profile at 350 µ m, 1.2 mm and 3.5 mm.The caveat of this method is that there is no assurance of finding a unique best fit model.However, the goal of this section is to prove that the observed properties of the continuumemission observed in the HH 80N core can be explained in terms of a protostar plus an in-falling envelope that is embedded inside the HH 80N core. Constraining a unique model witha flattened envelope and a disk requires additional mid-IR observations with high angularresolution and it is beyond the scope of this paper. We tested the logatropic density distribution by using 1080 different models withinthe following set of ranges for the external radius, mass infall rate and mass of the centralembedded object: 0.04 pc ≤ R ext ≤ × − M ⊙ yr − ≤ ˙ M i ≤ × − M ⊙ yr − and 0.5 M ⊙ ≤ M ∗ ≤ M ⊙ . In this space of parameters we expect to find meaningfulphysical solutions. To establish the goodness of the fit we calculated the χ as discussed inthe previous section. Figure 3 shows the best set of solutions for the χ estimated separatelyfor the SED and for the 1.2 mm and 350 µ m intensity distributions. We also calculated the χ function for the 3.5 mm band but the results are not included in the figure because no setof parameters can provide reasonable χ values at this band. We consider as good solutionsthose where the χ values of the intensity distributions and of the SED are all within the90% level of confidence.Figure 3 shows that there are no solutions that fit together the SED and the intensitydistribution of the 1.2 mm and 350 µ m maps (i.e., there is no overlap between the χ contoursof the maps and those of the SED). As an example, Figures 4 and 5 show, respectively,the predicted SED and intensity profiles (dash-dotted lines) of the SLS model that givesthe minimum χ for the SED, compared with the observed data. Except for the mid-IRwavelengths, the observed SED is reasonably well reproduced by the model. However, as 15 –Figure 5 shows, this model produces intensity profiles too flat and cannot reproduce theobserved intensity profiles in any of the bands. Given these discrepancies, we conclude thatin the logatropic case the mass is not distributed adequately in the envelope to match theobservations. We tested the isothermal density distribution by using 1800 different models within thefollowing set of ranges for external radius, mass infall rate and mass of the central embeddedobject: 0.03 pc ≤ R ext ≤ . × − M ⊙ yr − ≤ ˙ M i ≤ . × − M ⊙ yr − and 0.5 M ⊙ ≤ M ∗ ≤ M ⊙ . Figure 6 shows the best models from the χ analysis for the SIS case. Forthe same reasons as in the logatropic case, we have not included the 3.5 mm results in thefigure.The mass infall rates considered for the SIS collapse are higher than those consideredin the logatropic case. This is because, for a given value of the mass infall rate, the SISmodels yield less massive envelopes than the SLS models; so, in order to fit properly themm flux densities higher values of the mass infall rate are required in the SIS models (seeOsorio et al. 1999). For this reason, R ext becomes almost irrelevant and ˙ M i takes thedominant role in the fitting. Figure 6 (bottom panels) shows that for the M ∗ = 0 . M ⊙ case,there are solutions that apparently satisfy both the SED and (sub)mm intensity distributionconstraints (i.e., for 0.8 M ⊙ , the χ contours of the SED overlap those of the maps for R ext ≃ ˙ M i = (6.5-8.0) × − M ⊙ yr − ). Nevertheless, despite these solutions yieldingreasonable intensity distributions in the 1.2 mm and 350 µ m bands, they overestimate thetotal luminosity, specially at far-IR wavelengths. In Figure 4 (dashed line) we show anexample that illustrates this behavior (the predicted flux is almost one order of magnitudeabove the observed flux at 60 µ m). As in the SLS case, the modeled intensity profile at3.5 mm (dashed line in Fig. 5) is significantly weaker than the observed profile. These resultsare general and we can find solutions that can fit the single–dish intensity distribution of theenvelope but they predict an excess of luminosity and fail to reproduce the compact emissionseen in the 3.5 mm PdBI map. In conclusion, as in the SLS case, there is no SIS model thatcan fit the SED and the maps simultaneously. 16 – One of the caveats of the SLS and SIS models is that they are unable to fit the intensityof the compact source seen in the PdBI 3.5 mm map. This compact source could have a sig-nificant contribution from a circumstellar disk. In this section we model the source assuminga TSC envelope falling onto a disk surrounding the central object obtained from the catalogof D’Alessio et al. (2005). Apart from using a more realistic model, such a configuration isideal for two reasons. First, if the circumstellar disk is populated by millimeter-size grains,its emission can be significant at mm wavelengths and, because it is compact, it will be lessaffected by the interferometric filtering that affects the envelope emission. Second, becausethe SED at short wavelengths is very sensitive to the geometry of the source, the TSC en-velope and a disk with the proper inclination can provide the extinction required to fit themid-IR part of the SED, depending on the inclination of the axis of the system (envelopeplus disk). This would give luminosities similar to or below 110 L ⊙ . In addition, the infallluminosity is reduced because the material lands on a disk instead of falling directly ontothe protostar.We tested several TSC envelopes exploring values of the central luminosity < L ∗ < L ⊙ , outer radius of the envelope . < R ext < . pc, the radius of the expansion wave . < R ew < . pc, and reference density . × − < ρ < × − g cm − (thesevalues of ρ are equivalent to densities at 1000 AU between . × and . × cm − ,and correspond to values of the mass infall rate between . × − and . × − M ⊙ yr − for a 3 M ⊙ star). For the circumstellar disk we explored accretion rates from the disk toprotostar between − to − M ⊙ yr − , and we assume that the disk is irradiated with asimilar luminosity and has a similar inclination than the envelope. In Table 4 we give theparameters of our favored model. Figures 7 and 8 show the observed SED and intensityprofiles, respectively, predicted by our favored TSC model. These figures show that the SEDis well reproduced by the model in almost all the data points and that the modeled intensityprofiles fit reasonably well the observations within the calibration uncertainties.At 3.5 mm, the inclusion of a disk provides the flux needed to explain the observedintensity peak. We note that, possibly, a fine tuning of the mass accretion rate could providea more accurate fit for the intensity profiles. However, the range of accretion rates of theonline disk model grid is sampled in steps that vary one order of magnitude and the nextavailable model has an accretion rate too small and provides too faint mm emission. Never-theless, in this analysis we aimed to prove that our continuum observations can be explainedin the frame of standard star forming models and our favored model presented above fulfillsthis requirement. Table 5 shows a summary of the χ analysis (SED, single-dish and inter-ferometric maps) for this TSC (+disk) model (Figs. 7, 8), as well as for the SLS and SIS 17 –models shown in Figures 4, 5. The reduced χ values show that in the TSC case we obtaina better fit.
5. Discussion
The derived physical parameters of the selected TSC model listed in Table 4 suggestthat HH 80N-IRS1 is a very young Class 0 protostar. First, the predicted volume density forHH 80N-IRS1 at 1000 AU, n (H ) ≃ × cm − , is typical of Class 0 sources (Jørgensen etal. 2002), and greater than Class I sources (Jørgensen et al. 2002) and prestellar cores (Kirk,Ward-Thomson & André 2005; Tafalla et al. 2002); second, the estimated upper limit of themass infall rate, ∼ × − M ⊙ yr − (see Table 4), is compatible with the high values of themass infall rate typical of young Class 0 protostars (Maret et al. 2002), and yields to an ageof ∼ × yr for HH 80N-IRS1; finally, HH 80N-IRS1 fulfills the criteria proposed by Andréet al. (1993) for Class 0 objects, L submm /L bol & × − , which in our case is about 0.1. Onthe other hand, the derived luminosity of 105 L ⊙ for HH 80N-IRS1 is found in the thresholdbetween low mass and intermediate mass protostars. Given the large reservoir of mass of theHH 80N core (see 1.2 mm and 350 µ m map of Fig. 2) and the youth of the HH 80N-IRS1,we cannot discard further accumulation of material towards the central object.From the integrated emission of the residual map at 1.2 mm (resulting from the sub-traction of the emission of the synthetic map of the TSC model from the map observed withMAMBO) we can obtain a crude estimate of the mass of the HH 80N core outside the HH80N-IRS1 envelope. Assuming optically thin dust emission with β = 2 and a temperatureof ∼
14 K (the boundary temperature of the HH 80N-IRS1 envelope) we derive a mass of ∼ M ⊙ for this material. As we noted above, the HH 80N core, with a size of . × . pcand an estimated total mass of ∼ M ⊙ (20 M ⊙ of HH 80N-IRS1 + 10 M ⊙ of the rest ofthe HH 80N core), contains more material than the infalling envelope associated with IRS1.According to the results of our TSC modeling (see Table 4) the mass of the envelope is 20 M ⊙ , and the infall occurs within a radius of R ew = . × AU with the envelope beingstatic outside this radius. Furthermore, the molecular emission of tracers such CS, SO andHCO + extends over a region considerably larger than the HH 80N core, as traced by thedust continuum and ammonia line emissions. Indeed, the emission of these molecular tracershas been interpreted as arising from a contracting ring around HH 80N-IRS1 with an innerradius (the radius of the region where these molecular species appear to be depleted) of . × AU and an outer radius of × AU (Girart et al. 2001; Masqué et al. 2009). Theestimated average volume density of the molecular ring is in the × - . × cm − range(Masqué et al. 2009), which seems too high, given the estimated density in the static part 18 –of the HH 80N-IRS1 envelope ( ∼ × cm − , according to our modeling). Therefore, thekinematics and physical conditions in the molecular ring-like structure proposed by Girart etal. (2001) and Masqué et al. (2009) appear puzzling. The role of the HH 80/81/80N outflowin the properties of this molecular ring-like structure and the relationship with the onset ofthe star-forming process in the HH 80N core, in particular with the HH 80N-IRS1, protostaris an interesting issue that deserves further observational and theoretical investigation.
6. Conclusions
We have carried out dust continuum and ammonia line observations of the dense coreahead of HH 80N, complemented with archive data, covering a wide range of wavelengths.We analyzed the continuum data by means of self-consistent models using several approachesfor the envelope structure and we discuss the inclusion of a protostellar disk. Additionally,we compare ammonia observations of the (1,1) transition with continuum emission (andabsorption) maps. Our main conclusions are summarized as follows:1. The NH (1,1) emission shows a striking correlation with the dust continuum emissionand with the absorption silhouette seen in the 8 µ m Spitzer image. This indicatesthat the ammonia traces fairly well the distribution of gas and dust in the HH 80Ncore. Pending a proper analysis of the NH abundances, this preliminary assessmentpoints that there is no need to invoke photochemical effects caused by the nearby HH80N object to explain the distribution of ammonia in the HH 80N core. However, adetailed inspection of the ammonia map shows that an important part of the NH emission arises from the Southeastern part of the core, close to HH 80N, which couldbe due to a slight abundance enhancement.2. The continuum emission presents a peak at the same position ( α ( J h m . s , δ ( J − ◦ ′ . ′′ ) in all the bands (4.5 µ m, 8 µ m, 350 µ m, 1.2 mm and 3.5 mm).This emission peak is located at the center of the CO bipolar outflow found by Girartet al. (2001), suggesting the presence at this position of an embedded young stellarobject (HH 80N-IRS1) that powers the outflow.3. We find that the SED and the intensity distribution of the mm and submm emissionof HH 80N-IRS1 can be reproduced by a slowly rotating infalling envelope describedby the Terebey, Shu, and Cassen (TSC) solution, plus a circumstellar accretion disk.The mass of the envelope is 20 M ⊙ , the central luminosity is 105 L ⊙ , and the radiusof the infalling region is 1.5 × AU. The disk has a mass of 0.6 M ⊙ and a radius of300 AU. Such a configuration, together with the derived high values of the mass infall 19 –rate ( . × − ( M ∗ / M ⊙ ) / M ⊙ yr − ), and young age ( ∼ × yr), suggest theHH 80N-IRS1 may be a young Class 0 source.4. The APEX map at 350 µ m and, especially, the PdBI map at 3.5 mm, where theextended emission is resolved out, show signs of possible fragmentation suggestingthat other sources, in addition to HH 80N-IRS1, could be embedded inside the HH80N core. On the other hand, previous studies reveal that the molecular emission ofsome tracers is considerably more extended than the dust and NH emission presentedin this work. This suggests that the HH 80N core is surrounded by a larger molecularstructure whose properties could be influenced by the proximity of the HH 80/81/80Noutflow.G.A., R.E., J.M.G., J.M.M., and M.O. acknowledge support from MICINN (Spain)grant AYA2008-06189-C03 (co-funded with FEDER funds). G.A. and M.O. acknowledgepartial support from Consejería de Innovación, Ciencia y Empresa de la Junta de Andalucía(Spain). G.G. acknowledges support from CONICYT projects FONDAP No. 15010003 andBASAL PFB-06. We thank Susana Lizano for providing us the routines for the calculationof the logatropic density distribution. REFERENCES
André, P., Ward-Thompson, D., & Barsoni, M. 1993, ApJ, 406, 122Basu, S., & Mouschovias, T, Ch. 1994, ApJ, 432, 720Calvet, N., Hartmann, L., Kenyon, S. J.,& Whitney, B. A. 1994, ApJ, 434, 330Carey, S. J., Clark, F. O., Egan, M. P., Price, S. D., Shipman, R. F., & Kuchar, T. A. 1998,ApJ, 508, 721Carrasco-González, C., Rodríguez, L.F., Anglada, G., Martí, J., Torrelles, J.M., & Osorio,M., 2010, Science, 330, 1209Cassen, P., Mosman, A. 1981 Icarus, 48, 353D’Alessio, P., 1996, PhD. thesisD’Alessio, P., Merín, B., Calvet, N., Hartmann, L., & Montesinos, B. 2005, Rev. MexicanaAstron. Astrofis., 41, 61Davis, C. J., Dent, W. R. F, & Bell Burnell, S. J. 1990, MNRAS, 244, 173 20 –Fernández-López, M., Curiel, S., Girart, J. M., Ho, P. T. P., Patel, N., & Gómez, Y. 2011,AJ, 141, 72Girart, J. M., et al. 1994, ApJ, 435, L145Girart, J. M., Estalella, R. , & Ho, P. T. P. 1998, ApJ, 495, L59Girart, J. M., Estalella, R., Viti, S., Williams, D. A., & Ho, P. T. P. 2001, ApJ, 562, L91Girart, J. M., Viti, S., Williams, D. A., Estalella, R., & Ho, P. T. P. 2002, A&A, 388, 1004Girart, J. M., Viti, S., Estalella, & Williams, D. A. 2005, A&A, 439, 601Gyulbudaghian, A. L., Glushkov, Y. I., & Denisyuk, E. K. 1978, ApJ, 224, 137Jørgensen, J. K., Schöier, F. L., & van Dishoeck, E. F. 2002, A&A, 389, 908Kenyon, S. J., Calvet, N., & Hartmann, L. 1993, ApJ, 414, 676Kirk, J. M., Ward-Thomson, D., & André, P. 2005, MNRAS, 360, 1506Lizano, S., & Shu, F. H. 1989, ApJ, 342, 834Martí, J., Rodríguez, L.F., & Reipurth, B. 1993, ApJ, 416, 208Masqué, J. M., Girart, J. M., Beltrán, M. T., Estalella, R., & Viti, S. 2009, ApJ, 695, 1505Maret, S., Ceccarelli, C., Caux, E., Tielens, A. G. G. M., & Castets, A. 2002, A&A, 395,573McLaughlin, D. E., & Pudritz, R. E. 1996, AAS, 189, 1503Molinari, S., Noriega-Crespo, A., & Spinoglio, L. 2001, ApJ, 547, 292Morata, O., Girart J. M., & Estalella, R., 2003 A&A, 397, 181Morata, O., Girart J. M., & Estalella, R., 2005 A&A, 435, 113Osorio, M., Lizano, S., & D’Alessio, P. 1999, ApJ, 525, 808Osorio, M., D’Alessio, P., Muzerolle, J., Calvet, N., & Hartmann, L. 2003, ApJ, 586, 1148Osorio, M., Anglada, G., Lizano, S., & D’Alessio, P. 2009, ApJ, 649, 29Rathborne, J. M., Jackson, J. M., Chambers, E. T., Stojimirovic, I., Simon, R., Shipman,R., & Frieswijk, W. 2010, ApJ, 715, 310Rodríguez, L. F., Moran, J. M., Gottlieb, E. W., & Ho, P. T. P. 1980, ApJ, 235, 845Rudolph, A., & Welch, W. J. 1988, ApJ, 326, L31Schaller, G., Schaerer, D., Meynet, G., Maeder, A. 1992, A&AS, 96, 269Shu, F. H., Adams, F. C., & Lizano, S. 1987, ARA&A, 25, 23Tafalla, M., Myers, P. C., Caselli, P., Walmsley, C. M., & Comito, C. 2002, ApJ, 569, 815 21 –Taylor, S.D., & Williams, D.A. 1996, MNRAS, 282, 1343Terebey, S., Shu, F. H., & Cassen, P. 1984, ApJ, 286, 529Torrelles, J. M., Rodríguez, L. F., Cantó, J., Anglada, G., Gómez J. F., Curiel, S., & Ho, P.T. P. 1992, ApJ, 396, L95Torrelles, J. M., Gómez, J. F., Ho, P. T. P., Anglada, G., Rodríguez, L. F., & Cantó, J.1993, ApJ, 417, 655Viti, S., & Williams, D. A. 1999, MNRAS, 310, 517Viti, S., Girart, J. M., Garrod, R., Williams, D. A., & Estalella, R. 2003, A&A, 399, 187Viti, S., Girart, J. M., & Williams, D. A. 2006, A&A, 449, 1089Whyatt, W., Girart, J. M., Viti, S., Estalella, R., & Williams, D. A. 2010, A&A, 510, 74Wolfire, M. G., & Königl, A. 1993, ApJ, 415, 204
This preprint was prepared with the AAS L A TEX macros v5.2.
22 –Fig. 1.— Images of the HH 80N region at 4.5 µ m (top panel) and 8 µ m (middle panel),retrieved from the Spitzer archive, and 18.7 µ m (bottom panel) obtained with the VLT. Inthe middle panel, the dashed square shows the limits of the VISIR (VLT) field of view andthe dashed arrow gives approximately the orientation and extend of the outflow detected byGirart et al. 2001. 23 –Fig. 2.— Maps of the HH 80N region taken at 350 µ m (top panel), 1.2 mm (second panel),3.5 mm (third panel) and the velocity integrated (zero-order moment) of the main line ofNH (1,1) emission (bottom panel), superimposed on the Spitzer 8 µ m image. Contour levelsare 3, 4, 6, 9, 15, and 21 times 90 mJy beam − (350 µ m); − −
3, 3, 6, 9, 15, 21, 27, 39,and 50 times 1.5 mJy beam − (1.2 mm); − −
2, 2, 3, 5, 7 and 10 times 0.11 mJy beam − (3.5 mm); 3, 6, 10, 16 and 24 times 0.24 Jy beam − km s − (NH ). The beams are shown inthe upper left corner of the panels. The color scale of the infrared image has been modifiedwith respect to Fig. 1 in order to highlight the absorption feature (see text). 24 –Fig. 3.— Contour plots (solid line and gray scale) of the χ function derived for the fittingof the models of the collapse of the Singular Logatropic Sphere to the observed images. Rowscorrespond to the 350 µ m band (top), the 1.2 mm band (middle), and the sum of the χ function for the 350 µ m and 1.2 mm bands (bottom). The dotted contours shown in thebottom panels correspond to the χ function derived for the SED. Columns correspond to M ∗ = 0.5 M ⊙ (left), 0.7 M ⊙ (middle), and 0.9 M ⊙ (right). In all the cases the contour levelscorrespond to the confidence levels of 99%, 90% and 68% (1 σ ). These contours are relativeto the minimum χ value of each panel, which may vary significantly for different masses.For instance, in the bottom row, where we compare the results of the SED and single-dishmaps, the minimum χ values of the SED are 53.4, 33.2, and 19.2 for 0.5, 0.7, and 0.9 M ⊙ ,respectively; the sum of the minimum χ values of the 350 µ m and 1.2 mm maps are 102.2,119.0, and 122.2 for 0.5, 0.7, and 0.9 M ⊙ , respectively. 25 –Fig. 4.— Observed flux densities (dots; see Table 3) and predicted SEDs for HH 80N-IRS1,assuming models of the collapse of the Singular Logatropic Sphere (SLS; dash-dotted line)and of the Singular Isothermal Sphere (SIS; dashed line). Bars represent the uncertaintiesand arrows represent upper limits. The SLS model corresponds to the model that minimizesthe χ of the SED ( R ext = 0.1 pc, M ∗ = 0.7 M ⊙ , ˙ M i = 1.57 × − M ⊙ yr − , L bol = 69 L ⊙ , M env = 25 . M ⊙ , and β = 1.6). The SIS model corresponds to the model that minimizesthe χ function for the maps at 1.2 mm and 350 µ m ( R ext = 0.05 pc, M ∗ = 0.8 M ⊙ , ˙ M i = 1.05 × − M ⊙ yr − , L bol = 527 L ⊙ , M env = 4 . M ⊙ , and β = 1.1). 26 –Fig. 5.— Observed and modeled intensity profiles of HH 80N-IRS1 at 350 µ m (top), 1.2mm (middle) and 3 mm (bottom). The squares and triangles represent observed cuts alongthe major (P. A. ≃ ◦ ) and minor (P. A. ≃ ◦ ) axes of the HH 80N core, respectively.Bars represent the uncertainties. At 3.5 mm, as the source appears unresolved, we presentonly a cut along the major axis of the core (P. A. ≃ ◦ ). The dash-dotted lines correspondto cuts along the diameter of the synthetic maps of the Singular Logatropic Sphere modelthat minimizes the χ for the SED (see Fig. 4). The dashed lines correspond to cuts alongthe diameter of the synthetic maps of the Singular Isothermal Sphere model that minimizesthe χ for the maps (see Fig. 4). The dotted line represents the beam. 27 –Fig. 6.— Contour plots (solid line and gray scale) of the χ function derived for the fittingof the models of the collapse of the Singular Isothermal Sphere to the observed images of HH80N-IRS1. Rows correspond to the 350 µ m band (top), the 1.2 mm band (middle), and thesum of the χ function for the 350 µ m and 1.2 mm bands (bottom). The dotted contoursshown in the bottom panels correspond to the χ function derived for the SED. Columnscorrespond to M ∗ = 0.5 M ⊙ (left), 0.8 M ⊙ (middle), and 0.9 M ⊙ (right). In all the casesthe contour levels correspond to the confidence levels of 99%, 90% and 68% (1 σ ). Thesecontours are relative to the minimum χ value of each panel, which may vary significantlyfor different masses. For instance, in the bottom row, where we compare the results of theSED and single-dish maps, the minimum χ values of the SED are 47.1, 278.6, and 600.3for 0.5, 0.7, and 0.9 M ⊙ , respectively; the sum of the minimum χ values of the 350 µ m and1.2 mm maps are 40.6, 36.1, and 36.7 for 0.5, 0.7, and 0.9 M ⊙ , respectively. 28 –Fig. 7.— Observed flux densities (dots; see Table 3) and predicted SED for HH 80N-IRS1, assuming the Terebey, Shu, and Cassen (TSC) envelope model plus an accretiondisk. Bars represent the uncertainties and arrows represent upper limits. The dashed linerepresents the SED of our favored TSC envelope model ( R ext = 1 . × AU, M ∗ = 3 M ⊙ , ˙ M i ≃ . × − M ⊙ yr − , L bol = 105 L ⊙ , and M env = 20 M ⊙ ). The point dashed linerepresents the SED of the selected disk model ( ˙ M acc = 10 − M ⊙ yr − , R c =
300 AU, and i = 30 ◦ ). The solid line represents the resulting SED of out favored TSC envelope plus diskmodel. 29 –Fig. 8.— Observed and modeled intensity profiles of HH 80N-IRS1 at 350 µ m (top), 1.2mm (middle) and 3 mm (bottom). The squares and triangles represent observed cuts alongthe major (P. A. ∼ ◦ ) and minor (P. A. ∼ ◦ ) axes of the HH 80N core, respectively.Bars represent the uncertainties. At 3.5 mm, as the source appears unresolved, we presentonly a cut along the major axis of the core (P. A. ∼ ◦ ). The solid lines correspond to cutsalong the diameter of the synthetic maps of our favored Terebey, Shu, and Cassen envelopeplus disk model (see Fig. 7). The dotted line represents the beam. 30 –Table 1: Positions of the compact Spitzer sources a Peak PositionSource RA(J2000) DEC (J2000)HH 80N-IRS1 h m . s − ◦ ′ . ′′ HH 80N-IRS2 h m . s − ◦ ′ . ′′ HH 80N-IRS3 h m . s − ◦ ′ . ′′ a HH 80N-IRS1 position was derived from the 8 µ m image. HH 80N-IRS2 and HH 80N-IRS3 positions werederived from the 4.5 µ m image. Table 2: Source parameters derived from the PdBI 3.5 mm map a Peak Position I ν (peak) b S ν c Deconvolved Size P. A.Source RA(J2000) DEC (J2000) (mJy beam − ) (mJy) ( ′′ ) ( ◦ )HH 80N-IRS1 h m . s − ◦ ′ . ′′ . ± .
11 3 . ± .
19 5 . × . -17.4Southeastern Condensation h m . s − ◦ ′ . ′′ . ± .
06 2 . ± .
09 11 . × . -17.0 a Derived from a Gaussian fit with the task IMFIT of MIRIAD. b Peak Intensity. c Integrated flux density.
31 –Table 3: Summary of the continuum data of HH 80N-IRS1
Angular Aperture FluxWavelength Resolution a Size b Density c Observing( µ m) Instrument ( ′′ ) ( ′′ ) (Jy) Epoch Notes3500 IRAM PdBI 2.9 × ∼ × d ∼ × ∼
60 – ≤ ∼
55 – ≤ ×
300 – ≤
108 1983 Archive data60 IRAS 90 ×
282 – ≤ ×
276 – ≤ ∼ . ∼ e ×
270 – ≤ a For the mm and submm data the reported angular resolution corresponds to the FHWM of the beam size.For far-IR and mid-IR data it corresponds to the Point Spread Function. For Spitzer it corresponds to thepixel size, which is larger than the angular resolution. b The box for the 350 µ m and 1.2 mm data has a P.A. of 120 ◦ and it is chosen to include only HH 80N-IRS1(see § 3.1). The values given for IR data are the diameter of a circular aperture. c For the 350 µ m, 1.2 mm and 3.5 mm measurements we give a range in order to account for contaminationeffects from the Southeastern Condensation (see § 3.3). For the mid infrared points uncertainties are in-cluded into parenthesis. We adopted the IRAS and Akari flux values as upper limits because of possiblecontamination by background sources. d After smoothing with a Gaussian of FWHM = . ′′ . e Adopting a 5% of calibration uncertainty as indicated in the IRC Data User Manual.
32 –Table 4: Results of the TSC modelingEnvelope parameter Symbol ValueMass M M ⊙ Central luminosity L ∗ L ⊙ Radius of the expansion wave a R ew . × AUOuter radius R ext . × AUInclination angle i ◦ Centrifugal radius R c
300 AUReference density ρ × − gr cm − Density at r = 1000 AU n (1000 AU) . × cm − Mass infall rate b ˙ M i . × − M ⊙ yr − Disk parameter c Symbol ValueMass M disk M ⊙ Radius d R disk
300 AUInclination angle e i ◦ Mass accretion rate ˙ M acc − M ⊙ yr − Viscosity parametrization α a min µ mMaximum grain size a max a Radius of the infalling region. Outside this radius the envelope remains static. b Obtained adopting a mass of M ∗ = 3 M ⊙ for the embedded protostar (i.e. the infall rate value is an upperlimit). c Obtained assuming that the disk is irradiated by a luminosity equal to the central luminosity L ∗ = 105 L ⊙ . d Assumed to coincide with the centrifugal radius derived for the envelope. e Assumed to coincide with the inclination angle derived for the envelope.
33 –Table 5. Reduced χ ResultsSingle-dish InterferometricSED maps a map b SLS model 34.1 31.5 68.3SIS model 441.4 3.3 13.3TSC+disk model 3.5 8.1 5.5 a µ m map obtained with LABOCA at APEX12 m telescope. bb