Millimeter interferometric observations of FU Orionis-type objects in Cygnus
aa r X i v : . [ a s t r o - ph . S R ] S e p Astronomy&Astrophysicsmanuscript no. paper c (cid:13)
ESO 2018November 21, 2018
Millimeter interferometric observations ofFU Orionis-type objects in Cygnus ´A. K ´osp´al Leiden Observatory, Leiden University, PO Box 9513, 2300 RA Leiden, The Netherlandse-mail: [email protected]
Received date; accepted date
ABSTRACT
Context.
FU Orionis-type objects (FUors) are low-mass young eruptive stars that probably represent an evolutionary phase charac-terized by episodic periods of increased accretion rate from the circumstellar disk to the star. Theory predicts that a circumstellarenvelope, the source of continuous mass infall onto the disk, is necessary for triggering such accretion bursts.
Aims.
We intend to study the spatial and velocity structure of circumstellar envelopes around FUors by means of molecular lineobservations at millimeter wavelengths. We target three prototypical FUors as well as an object possibly in a pre-outburst state.
Methods.
We present archival PdBI interferometric observations of the J = CO at 110.2 GHz. For three of our targets,these represent the first millimeter interferometric observations. The data allow the study of the molecular environment of the objectson a spatial resolution of a thousand AU and a velocity resolution of 0.2 km s − . Results.
Strong, narrow CO(1–0) line emission is detected from all sources. The emission is spatially resolved in all cases, withdeconvolved sizes of a few thousand AUs. For V1057 Cyg and V1331 Cyg, the emitting area is rather compact, suggesting that theorigin of the emission is a circumstellar envelope surrounding the central star. For V1735 Cyg, the CO emission is o ff set from thestellar position, indicating that the source of this emission may be a small foreground cloud, also responsible for the high reddeningof the central star. The CO emission towards V1515 Cyg is the most extended in the sample, and apparently coincides with thering-like optical reflection nebula associated with V1515 Cyg.
Conclusions.
We suggest that millimeter interferometric observations are indispensable for a complete understanding of the circum-stellar environment of FUors. Any theory of the FUor phenomenon that interprets the geometry of the circumstellar structure and itsevolution using single beam measurements must be checked and compared to interferometric observations in the future.
Key words. stars: formation – infrared: stars – circumstellar matter – stars: individual: V1057 Cyg, V1331 Cyg, V1515 Cyg,V1735 Cyg
1. Introduction
FU Orionis-type objects, or shortly FUors, constitute asmall group of young stars characterized by large out-bursts in visible light, attributed to highly enhanced accretion(Hartmann & Kenyon, 1996). During these outbursts, accretionrates from the circumstellar disk to the star are in the order of10 − M ⊙ / yr, three orders of magnitude higher than in quiescence.Enhanced accretion is often accompanied by enhanced massloss: most FUors have optical jets, molecular outflows, and op-tically visible ring-like structures that are sometimes explainedby expanding shells thrown o ff during previous outbursts.During a single, century-long outburst, as much as 0.01 M ⊙ of material can be dumped onto the stellar surface. Thus, the in-ner disk needs to be replenished after each outburst, possibly bymaterial from the infalling envelope. Recent theoretical studiesshow that the continuous infall from the envelope is also nec-essary to trigger the outbursts (Vorobyov & Basu, 2006). Aftermany outbursts, the envelope vanishes, and the object finally en-ters a state of permanently low accretion. This general paradigmof the evolution of young, low-mass stars was invoked byQuanz et al. (2007) to explain the observed diversity of FUors:while some objects are still deeply embedded (e.g. L1551 IRS 5,V1735 Cyg), others have already cleared away part of their en-velopes (e.g. V1057 Cyg, V1515 Cyg). Statistics show that prob-ably all low-mass young stars undergo FUor-like phases during their evolution, implying that FUors might be the clue objects tostudy envelope evolution and dispersal.Interferometric observations of molecular line emission havebeen successfully used to probe small-scale structure of molec-ular material around young stellar objects. For example, fora sample of low-mass YSOs in Taurus (Hogerheijde et al.,1998), Serpens (Hogerheijde et al., 1999), and Ophiuchus(van Kempen et al., 2009), it was found that most YSOs are sur-rounded by compact envelopes of up to a few thousand AU inradius, well traced by CO and C O. Condensations, inho-mogeneities in the envelopes can be seen in HCO + and CO.The HCO + and HCN molecules trace the walls of the outflow,while SiO and SO emission originates from shocked materialin the outflow. These observations provided detailed kinemati-cal picture of the envelopes (for molecular studies of some in-dividual young stellar objects, see e.g. Jørgensen et al. 2004;Matthews et al. 2006; Brinch et al. 2009).Despite being “accretion laboratories”, there are relativelyfew molecular gas observations published for FUors. Single-dishobservations often show the presence of strong CO emission to-wards FUors, and line profiles sometimes indicate the presenceof molecular outflows (e.g. Hartmann & Kenyon 1996, and ref-erences therein). However, due to the large single dish beam, it isnot possible to study the spatial distribution of the emission us-ing these data. Out of the 9 FUors listed in Hartmann & Kenyon (1996), millimeter interferometric observations are only avail-able for L1551 IRS 5 (e.g. Momose et al., 1998).In this paper, we present interferometric observations of themolecular emission from four FUor-type objects in Cygnus:V1057 Cyg, V1331 Cyg, V1515 Cyg, and V1735 Cyg. With theexception of V1331 Cyg (Levreault, 1988; McMuldroch et al.,1993), these data represent the first millimeter interferometricobservations for our targets. We use the CO(1–0) line emis-sion to perform a high spatial and spectral resolution study of thegaseous material around FUors. We analyze the spatial and kine-matic structure of the CO gas, and check whether the emissioncan be associated with the circumstellar envelopes. Our resultscan be compared with those obtained for normal YSOs. Such acomparison may also contribute to the on-going debate whetherall young stars undergo an eruptive phase in their early evolutionor FUors are atypical objects.
2. Observations
We reduced unpublished CO observations of our targets ob-tained with the Plateau de Bure Interferometer (PdBI) on May30 and 31, 1993 (program ID: C057, PI: D. Fiebig). The ob-servations were carried out in snapshot mode (about 1 houron-source correlation time per object), using four antennas(4D1 configuration), and the baselines ranged from 24 to 64 m.The receiver was tuned to the CO(1–0) line at 110.2 GHz(lower sideband), the channel spacing was 78 kHz (0.21 km / s).At this wavelength, the single dish HPBW is 45 ′′ . Brightquasars (3C454.3, 2005 + + + h m . s + ◦ ′ . ′′ h m . s + ◦ ′ . ′′ h m . s + ◦ ′ . ′′ h m . s + ◦ ′ . ′′ . The rms phase noise was less than 18 ◦ for allobservations. We estimate a flux calibration accuracy of ≈ ′′ × ′′ , which were then deconvolved usingthe Clark CLEAN method. The synthesized beam is approxi-mately 7 ′′ × ′′ while the rms noise in the channel maps is ≈ / beam. For the exact phase centers and beam parameters, seethe log of observations in Table 1.
3. Results
The CO(1–0) spectra of our sources, integrated over thewhole 64 ′′ × ′′ maps, are plotted in Fig. 1. Strong CO(1–0) emission was detected towards all sources with high signal-to-noise ratio. In all cases, we found an emission line coincid-ing with the systemic velocity of each source (from optical andnear-IR spectra, Herbig 1977, Kenyon & Hartmann 1989, andChavarr´ıa et al. 1979 determined heliocentric radial velocities of − ± − ± − ±
2, and − ± − , coresponding toLSR velocities of 1.9, 0.6, 5.2, and 3.9 km s − for V1057 Cyg,V1331 Cyg, V1515 Cyg, and V1735 Cyg, respectively). We col-lapsed a few channels centered on the peak emission, and cal-culated visibility amplitudes as a function of uv radius (Fig. 2),and produced velocity-integrated maps (Fig. 3). The LSR veloc-ities where the emission peaks are listed in Table 1. Velocity- http: // / IRAMFR / GILDAS -20 -10 0 10 20 30Velocity (km s -1 )-10123 I n t e g r a t e d f l u x ( J y ) V1057 Cyg -20 -10 0 10 20Velocity (km s -1 )-1012345 I n t e g r a t e d f l u x ( J y ) V1331 Cyg-20 -10 0 10 20 30Velocity (km s -1 )-202468 I n t e g r a t e d f l u x ( J y ) V1515 Cyg -10 0 10 20 30Velocity (km s -1 )-10123 I n t e g r a t e d f l u x ( J y ) V1735 Cyg
Fig. 1. CO(1–0) spectra of our targets obtained with the PdBI.Fluxes were integrated over an area of 64 ′′ × ′′ centered on thephase center.integrated line fluxes calculated for the emission areas visible inFig. 3 are also listed in Table 1. The line profile of V1057 Cyg is the broadest in our sam-ple (FWHM ≈ − ), and is centered at 4.6 km s − . Thisis consistent with the velocity seen in single dish CO(1–0), CO(2–1), and CO(1–0) data by Bechis & Lo (1975)and Levreault (1988). V1331 Cyg shows a narrow CO(1–0)line (FWHM ≈ − ). McMuldroch et al. (1993) presentsOwens Valley interferometric observations of V1331 Cyg inthe same line, as well as in other transitions and other iso-topes of CO. The velocities of these lines are consistent withours. V1515 Cyg shows two separate, narrow emission com-ponents (FWHM ≈ − ) at 5.1 km s − and at 11.9 km s − .Both of these components are also visible in single dish CO(3–2) and CO(2–1) data from Evans et al. (1994). The spectrumof V1735 Cyg shows a narrow, single-peaked CO(1–0) line(FWHM ≈ − ). The shape and position of this line co-incides well with that of the CO(2–1) line detected in singledish data by Evans et al. (1994). The CO(2–1) line from thesame paper shows self-absorption at this velocity, and broad linewings indicating outflow activity. We plotted position-velocitydiagrams along di ff erent angles through our sources, but exceptfor V1515 Cyg (which will be discussed later), we found no sig-nificant velocity gradients. In Fig. 2 we plotted the visibility amplitudes as a function of uv radius. Since these are snapshot observations, the uv -space isonly partially sampled; there are ranges of uv radii where no vis-ibility information is available. However, even with this coarsesampling, the graphs show a decline of amplitude with uv ra-dius, indicating that our sources are spatially resolved. Simple2D Gaussian fits to the maps presented in Fig. 3 also indicate α δ FWHM P.A. v LSR
Flux(pc) ( ′′ × ′′ ) ( ◦ ) (h:m:s) ( ◦ : ′ : ′′ ) ( ′′ × ′′ ) ( ◦ ) (km s − ) (Jy km s − )V1057 Cyg 600 1993-May-31 7.02 × + × × + × − × + × × + × Table 1.
Summary of PdBI observations. Distances are from Sandell & Weintraub (2001). The FWHM of the fitted Gaussians aredeconvolved sizes.
10 20 30 40 50 60UV radius (m)01234 A m p li t u d e ( J y ) V1057 Cyg 10 20 30 40 50 60UV radius (m)01234 A m p li t u d e ( J y ) V1331 Cyg10 20 30 40 50 60UV radius (m)01234 A m p li t u d e ( J y ) V1515 Cyg 10 20 30 40 50 60UV radius (m)01234 A m p li t u d e ( J y ) V1735 Cyg
Fig. 2.
Visibility amplitudes as a function of uv radius. Data werebinned in 1 m-wide bins and the error bars indicate the dispersionof data points within one bin.that our targets are spatially resolved. The deconvolved sizes andPAs of the Gaussian fits are listed in Table 1. Gaussian modelsfitted to the visibilities give sizes within ± ′′ and PAs within ± ◦ to those determined from the Gaussian fits to the images.Thus, we consider these numbers as representative uncertaintiesof the values given in Table 1. We emphasize that these fits arenot physical envelope models, and thus the obtained parametersshould be regarded only as rough quantitative estimates of thesize of the emitting region.V1057 Cyg and V1331 Cyg are the most compact sources inour sample, with FWHM of about 10 ′′ × ′′ . Moreover, in bothcases, the emission in centered precisely on the optical positionof the stars. V1515 Cyg shows the most extended emission in oursample. The stellar position seems to be located at the southerntip of a bright, slightly curved, elongated filament, but extendedemission can be seen both to the south and to the north of thestar ( ≈ ′′ across). A map produced for the channels around theemission line at 11.9 km s − reveals that this separate velocitycomponent comes from a compact area ≈ ′′ to the northwest(P.A. 330 ◦ ) of the optical source. The bulk of the emission in theV1735 Cyg region comes from an area with a FWHM of ≈ ′′ centered 2 . ′′ ′′ from the stellar position. s s s s s h m s Right Ascension 15’ 0"15’ 10"15’ 20"15’ 30"15’ 40"44
15’ 50" D e c li n a t i o n V1057 Cyg 3.19 ... 6.59 km s -1 s s s s s h m s Right Ascension 21’ 20"21’ 30"21’ 40"21’ 50"22’ 00"50
22’ 10" D e c li n a t i o n V1331 Cyg -1.17 ... 0.74 km s -1 s s s s h m s Right Ascension 12’ 0"12’ 10"12’ 20"12’ 30"12’ 40"42
12’ 50" D e c li n a t i o n V1515 Cyg 4.47 ... 5.74 km s -1 s s s s s h m s Right Ascension 31’ 40"31’ 50"32’ 00"32’ 10"32’ 20"47
32’ 30" D e c li n a t i o n V1735 Cyg 3.98 ... 5.47 km s -1 Fig. 3. CO(1–0) maps of our targets, integrated in the velocityranges indicated above the images. White asterisks indicate op-tical positions from SIMBAD. Beam sizes are ≈ × ′′ . The noiselevel is σ = − ; the solid contours are 2 σ , 4 σ , 6 σ , ...;the dashed contours are 0, − σ , − σ . Following Scoville et al. (1986), the total H mass can be calcu-lated from the observed CO(1–0) line fluxes by M H = . × − × ( T X + . e − . / T X τ CO − e − τ D X ( CO) Z S ν dv M ⊙ , where T X is the excitation temperature, τ is the optical depth, D is the distance of the source, X ( CO) is the molecularabundance relative to H , and R S ν dv is the velocity-integratedline flux in units of Jy km s − . Using T X =
50 K (as determinedfrom dust continuum observations by Sandell & Weintraub2001), assuming the lines to be optically thin ( τ ≪ X ( CO) = × − (Langer & Penzias, 1993), the totalgas masses are 0.12, 0.06, 0.43, and 0.10 M ⊙ for V1057 Cyg,V1331 Cyg, V1515 Cyg, and V1735 Cyg, respectively. Ofcourse these numbers should be considered lower limits if thelines are actually optically thick, or if part of the CO emis-sion is resolved out by the interferometer. However, these val-ues are in good agreement with total masses derived fromsingle-dish 850 µ m dust continuum maps of comparable sizeby Sandell & Weintraub (2001), which were 0.10, 0.13, 0.15,and 0.42 M ⊙ for V1057 Cyg, V1331 Cyg, V1515 Cyg, and V1735 Cyg, respectively. This suggests that our interferometricobservations probably recover most of the CO emission.
Although the main focus of our data analysis is the study of the CO emission, the PdBI observations could also be used to lookfor 2.7 mm continuum emission by excluding channels aroundthe lines and collapsing the remaining channels. The resultingcontinuum maps have a typical rms noise of 2 mJy / beam. Wedetected 2.7 mm continuum emission only for one of our tar-gets, V1331 Cyg. This is not surprising because, based on theobservations of Sandell & Weintraub (2001), out of our four tar-gets, V1331 Cyg is the brightest also at 1.3 mm and at 850 µ m.The 2.7 mm emission we detected is not resolved, its positionis consistent with the optical stellar position, and the measuredtotal flux is 12 ± κ . = g − (Ossenkopf & Henning, 1994),an emissivity law of κ ∼ λ − and dust temperature of 50 K(Sandell & Weintraub, 2001), this flux corresponds to a total(gas + dust) mass of 0.19 M ⊙ , similar to what we obtained fromthe line flux of V1331 Cyg. Using 6 mJy as a 3 σ upper limitfor the 2.7 mm continuum flux of the other sources, we cangive upper limits of 0.11, 0.32, and 0.26 M ⊙ for V1057 Cyg,V1515 Cyg, and V1735 Cyg, respectively.
4. Discussion
In the following subsections, we discuss what our interferomet-ric CO(1–0) data can add to what we already know about thecircumstellar environment of our targets based on optical im-ages, sub-millimeter continuum maps, CO line observations, andSED analysis from the literature. We will also discuss whetherthe detected CO emission can be associated with the en-velopes.
After its outburst in 1969-70, an eccentric ring-like nebulosity of1 ′ × . ′ CO data, Levreault (1988) andEvans et al. (1994) concluded the presence of a molecular out-flow on a similar spatial scale. The surroundings of V1057 Cygwas mapped by Sandell & Weintraub (2001), whose 850 µ mcontinuum image indicates the presence of a rather compact butresolved ( ≈ ′′ ) source coinciding with the star and a fainter,north-south oriented filament.In our CO(1–0) data we do not detect emission either fromthe ring-like reflection nebula of from the north-south filament.However, our interferometric observations definitely resolve thecentral source: the deconvolved Gaussian FWHM at the distanceof V1057 Cyg corresponds to about 5800 × µ msilicate emission feature implies a close to pole-on geometry forthe source, in accordance with its relatively symmetric shapein our interferometric observations. The circumstellar mass of0.12 M ⊙ derived from our CO observations, consistently withthe dust mass estimate by Sandell & Weintraub (2001), is ratherhigh, significantly exceeding the typical disk masses of T Tauri-type stars.
V1331 Cyg is not a FUor in its present state, but – based onthe similarity of its spectrum to that of V1057 Cyg prior tomaximum light – is probably in a pre-outburst state, or be-tween outbursts (McMuldroch et al., 1993). Single-dish and in-terferometric CO and CO observations by Levreault (1988)and McMuldroch et al. (1993) revealed a complex circumstel-lar environment containing a molecular outflow approximatelyalong the line of sight, a flattened gaseous envelope of about6000 × ×
28 000 AU. The latter coincides well with thelarge ring-shaped optical reflection nebula seen by Quanz et al.(2007).Our interferometric CO observations of V1331 Cyg, con-sistently with the similar angular resolution measurements ofMcMuldroch et al. (1993), reveal a rather compact structure to-wards the star with a deconvolved Gaussian FWHM of about5300 × µ m maps ofSandell & Weintraub (2001), although it is only marginally re-solved ( . ′′ or 3300 AU). We suggest that both the CO gasemission and the dust thermal emission is originated in a cir-cumstellar envelope, probably the same flattened structure thatwas proposed by McMuldroch et al. (1993) to explain the natureof the CO emission. It is probable that the dust in this flattenedenvelope scatters the stellar light and gives rise to the inner ringobserved in optical images by Quanz et al. (2007).The presence of a circumstellar envelope around V1331 Cygis also consistent with the SED, which exhibits significant in-frared excess indicating substantial amounts of circumstellarmaterial (e.g. ´Abrah´am et al., 2004). The geometry proposed byMcMuldroch et al. (1993) and by Quanz et al. (2007) suggeststhat we see the inner part of the system through a conical cav-ity filled with a pole-on molecular outflow. However, we detect CO emission neither from the outflow nor from the outer, largeexpanding ring, in agreement with the interferometric observa-tions of McMuldroch et al. (1993).
V1551 Cyg brightened in optical light slowly during the 1940sand 1950s (Herbig, 1977). Photographic plates from this timeshow a bright, narrow arc of nebulosity extending to the northand west of the star. Later images from the 1970s to the presentday still show this northern arc, but a brighter nebulosity is alsovisible to the south and west, together forming a nearly com-plete circular ring with a diameter of ≈ ′′ (corresponding to s s s s h m s Right Ascension 12’ 0"12’ 10"12’ 20"12’ 30"12’ 40"42
12’ 50" D e c li n a t i o n Fig. 4.
The optical and millimeter environment of V1515 Cyg.The grayscale image in the background is an SDSS r-band im-age; the gray contours are the same CO(1–0) contours as inFig. 3. White asterisk indicate the optical stellar position.16 000 AU, see also the Sloan Digital Sky Survey r-band imagefrom 2003 in Fig. 4). The fact that the size of this ring did notchange significantly for 60 years but the brightness distributiondid, suggests that it is a reflection nebula similar in nature to thataround V1057 Cyg.Our CO(1–0) measurement shows arc-shaped emission,which, plotted over the optical image in Fig. 4, clearly coin-cides with the ring-shaped reflection nebula. The similar mor-phology of the CO emission and the scattered light, as well asthe matching radial velocities of the CO emission and the starsuggest that the detected molecular gas is physically associatedwith the FUor. Similarly to V1057 Cyg, the CO line profileof V1515 Cyg in Evans et al. (1994) indicate the presence of amolecular outflow. Our CO(1–0) data, however, show no high-velocity line wings; the emission line is narrow and thus proba-bly traces quiescent, not outflowing material. We detect a small( . − ) velocity di ff erence between the northern and thesouthern part of the ring, the southern part being slightly morered-shifted. However, the ring is clearly not expanding.A map of the 850 µ m dust continuum fromSandell & Weintraub (2001) shows faint extended emis-sion towards V1515 Cyg and also at about 10 ′′ to the northand northwest to the star. Although these observations weremade with a relatively large (15 ′′ ) beam, the results are notinconsistent with a potentially arc-shaped dust emission. It isthus possible that the ring of material that is responsible for theoptical reflection nebula and for the CO emission also emitsin dust continuum.Although the observed molecular emission does not exactlypeak towards the star, some dust and gas must be located in thevicinity of V1515 Cyg forming a circumstellar envelope. Thisclaim is suggested by the SED of the FUor, which was mod-eled with a disk + envelope geometry by Turner et al. (1997) andGreen et al. (2006). However, this envelope does not show upas a separate localized peak in our CO map. We measured thehighest gas mass towards this object within our sample, which can probably be explained by the contribution from both the cir-cumstellar envelope and from the arc-shaped feature. Thus the0.43 M ⊙ is an upper limit for the envelope mass. V1735 Cyg had its outburst some time in the 1950s or the 1960s(Elias, 1978). The star is associated with faint patches and fila-ments of reflected light within 1 ′ . Submillimeter dust continuumobservations by Sandell & Weintraub (2001) indicate the pres-ence of two sources: one associated with (albeit slightly o ff setfrom the optical position of) V1735 Cyg, and a brighter one lo-cated about 20-24 ′′ to the northeast (the deeply embedded Class Iprotostar V1735 Cyg SM1). Single-dish CO and CO obser-vations (Levreault, 1983; Richardson et al., 1985; Evans et al.,1994) revealed a complex molecular gas structure in the vicin-ity of these sources. The CO emission is extended in the ar-cminute spatial scale, and broad line wings indicate outflow ac-tivity. Evans et al. (1994) concluded that both V1735 Cyg andSM1 drive molecular outflows.Similarly to the other three FUors in our sample, our COinterferometric observations of V1735 Cyg show narrow emis-sion with no line wings. This again points to quiescent gas insome kind of envelope. Our map shows a compact peak closeto V1735 Cyg, but with a definite o ff set of 2 . ′′ ff set from the star, as long as itcauses enough interstellar extinction to explain the observed col-ors. Based on the fact that the CO emission we detect is o ff setfrom the stellar position, we propose that a significant part of theemission is coming from this foreground dark cloud. This sug-gestion is supported by the results of Quanz et al. (2007), who,based on the analysis of mid-infrared ice features, concludedthat the extinction for V1735 Cyg might be caused by ices some-where in the line of sight to the source, rather than material re-lated to the young star. The foreground structure, however, maynot be completely unrelated to the star because of their identicalradial velocities.For continuum observations such as those presented inWeintraub et al. (1991) and in Sandell & Weintraub (2001), thedominant source is not the FUor, but V1735 Cyg SM1. Thissource also emits molecular line emission, as evidenced by thesingle-dish CO(3–2) map presented by Evans et al. (1994).Interestingly, we did not detect V1735 Cyg SM1 in CO. Thereason for the non-detection may be partly due to the fact thatV1735 Cyg SM1 is separated from V1735 Cyg by about 20 ′′ ,thus it is at the very edge of our primary beam. Additionally,the submillimeter source is very extended, thus the interferome-ter may filter out most of its emission.
5. Summary and Conclusions
In this paper, we present interferometric observations of the CO(1–0) line of four well known FUors, young stellar objectscharacterized by large optical outbursts due to enhanced disk ac-cretion. For V1057 Cyg, V1515 Cyg, and V1735 Cyg, these rep-resent the first millimeter interferometric data published so far.This makes it possible to study the gas distribution on a few thou-sand AU spatial scale.Although all of our sources are known to drive molecu-lar outflows, as evidenced by the high-velocity line wings of CO, our data suggest that the CO(1–0) emission tracesquiescent gas. With the exception of V1515 Cyg, the size ofthe emitting region is within a few thousand AU, consistentwith typical circumstellar envelope sizes. Gas masses calculatedfrom our CO line fluxes and from dust continuum data fromSandell & Weintraub (2001) are also consistent. This indicatesthat the CO emission seen towards our sources is mostly origi-nated from a relatively compact circumstellar envelope, or, as inthe case of V1735 Cyg, possibly from a small foreground cloud.All of our sources are surrounded by reflection nebulosities,which are probably pre-existent structures illuminated by thebrightened central source. With the exception of V1515 Cyg, wedetect no CO emission from these structures. For V1515 Cyg,the CO emission coincides with the ring-shaped optical reflec-tion nebula. This indicates that the source is surrounded by a ringof material that on one hand scatters the optical light of the cen-tral star, and on the other hand emits at millimeter wavelengths.An important consequence is that attributing unresolved, single-dish millimeter fluxes of V1515 Cyg to the circumstellar enve-lope may not be entirely correct.Based on the appearance of the 10 µ m silicate feature,Quanz et al. (2007) defined two categories of FUors. They arguethat objects showing the feature in absorption (e.g. V1735 Cyg)are younger, still embedded in a circumstellar envelope.Objects showing the silicate band in emission (e.g. V1057 Cyg,V1331 Cyg, and V1515 Cyg) are more evolved, with direct viewon the surface layer of the accretion disk. However, in caseof V1735 Cyg, the millimeter emission and the absorption atshorter wavelengths may be due to a foreground cloud, andconsequently the object may be more evolved (less embed-ded) than it appears. Our CO data, especially the case ofV1515 Cyg and V1735 Cyg, demonstrate that millimeter emis-sion and consequently absorption at other wavelengths may notbe necessarily or exclusively associated with circumstellar en-velopes. Thus, large beam, single-dish data alone are probablynot enough to obtain a complete picture of the circumstellar en-vironment of FUors. We suggest that any theory of the FUor phe-nomenon that interprets the geometry of the circumstellar mate-rial and its evolution using single beam measurements must bechecked and compared to interferometric observations in the fu-ture. Millimeter observations with facilities such as ALMA forboth the dust continuum and for the line emission of moleculessuch as CO, C O, HCO + , HCN, etc. will be a fruitful directionin future studies of FUors. Acknowledgements.
The author thanks Prof. Wolfgang J. Duschl for mak-ing available the PdBI data and Dr. Roberto Neri for his help during thedata reduction. Discussions with Dr. Michiel Hogerheijde and Dr. MariaKun greatly improved the presentation of the data analysis. Funding for theSloan Digital Sky Survey (SDSS) and SDSS-II has been provided by theAlfred P. Sloan Foundation, the Participating Institutions, the National ScienceFoundation, the U.S. Department of Energy, the National Aeronautics andSpace Administration, the Japanese Monbukagakusho, and the Max PlanckSociety, and the Higher Education Funding Council for England. The SDSSWeb site is http: // / . The SDSS is managed by the AstrophysicalResearch Consortium (ARC) for the Participating Institutions. The ParticipatingInstitutions are the American Museum of Natural History, Astrophysical InstitutePotsdam, University of Basel, University of Cambridge, Case Western ReserveUniversity, The University of Chicago, Drexel University, Fermilab, the Institutefor Advanced Study, the Japan Participation Group, The Johns HopkinsUniversity, the Joint Institute for Nuclear Astrophysics, the Kavli Institutefor Particle Astrophysics and Cosmology, the Korean Scientist Group, theChinese Academy of Sciences (LAMOST), Los Alamos National Laboratory,the Max-Planck-Institute for Astronomy (MPIA), the Max-Planck-Institute forAstrophysics (MPA), New Mexico State University, Ohio State University,University of Pittsburgh, University of Portsmouth, Princeton University, theUnited States Naval Observatory, and the University of Washington. References ´Abrah´am, P., K´osp´al, ´A., Csizmadia, S., et al. 2004, A&A, 428, 89Bechis, K. P. & Lo, K. Y. 1975, ApJ, 201, 118Brinch, C., Jørgensen, J. K., & Hogerheijde, M. R. 2009, A&A, 502, 199Chavarr´ıa, C., Appenzeller, I., & Bertout, C. 1979, A&AS, 36, 465Elias, J. H. 1978, ApJ, 223, 859Evans, II, N. J., Balkum, S., Levreault, R. M., Hartmann, L., & Kenyon, S. 1994,ApJ, 424, 793Goodrich, R. W. 1987, PASP, 99, 116Green, J. D., Hartmann, L., Calvet, N., et al. 2006, ApJ, 648, 1099Hartmann, L. & Kenyon, S. J. 1996, ARA&A, 34, 207Herbig, G. H. 1977, ApJ, 217, 693Hogerheijde, M. R., van Dishoeck, E. F., Blake, G. A., & van Langevelde, H. J.1998, ApJ, 502, 315Hogerheijde, M. R., van Dishoeck, E. F., Salverda, J. M., & Blake, G. A. 1999,ApJ, 513, 350Jørgensen, J. K., Hogerheijde, M. R., van Dishoeck, E. F., Blake, G. A., &Sch¨oier, F. L. 2004, A&A, 413, 993Kenyon, S. J. & Hartmann, L. 1989, ApJ, 342, 1134Langer, W. D. & Penzias, A. A. 1993, ApJ, 408, 539Levreault, R. M. 1983, ApJ, 265, 855Levreault, R. M. 1988, ApJS, 67, 283Matthews, B. C., Hogerheijde, M. R., Jørgensen, J. K., & Bergin, E. A. 2006,ApJ, 652, 1374McMuldroch, S., Sargent, A. I., & Blake, G. A. 1993, AJ, 106, 2477Momose, M., Ohashi, N., Kawabe, R., Nakano, T., & Hayashi, M. 1998, ApJ,504, 314Ossenkopf, V. & Henning, T. 1994, A&A, 291, 943Quanz, S. P., Henning, T., Bouwman, J., et al. 2007, ApJ, 668, 359Richardson, K. J., White, G. J., Avery, L. W., Lesurf, J. C. G., & Harten, R. H.1985, ApJ, 290, 637Sandell, G. & Weintraub, D. A. 2001, ApJS, 134, 115Scoville, N. Z., Sargent, A. I., Sanders, D. B., et al. 1986, ApJ, 303, 416Turner, N. J. J., Bodenheimer, P., & Bell, K. R. 1997, ApJ, 480, 754van Kempen, T. A., van Dishoeck, E. F., Salter, D. M., et al. 2009, A&A, 498,167Vorobyov, E. I. & Basu, S. 2006, ApJ, 650, 956Weintraub, D. A., Sandell, G., & Duncan, W. D. 1991, ApJ, 382, 270´Abrah´am, P., K´osp´al, ´A., Csizmadia, S., et al. 2004, A&A, 428, 89Bechis, K. P. & Lo, K. Y. 1975, ApJ, 201, 118Brinch, C., Jørgensen, J. K., & Hogerheijde, M. R. 2009, A&A, 502, 199Chavarr´ıa, C., Appenzeller, I., & Bertout, C. 1979, A&AS, 36, 465Elias, J. H. 1978, ApJ, 223, 859Evans, II, N. J., Balkum, S., Levreault, R. M., Hartmann, L., & Kenyon, S. 1994,ApJ, 424, 793Goodrich, R. W. 1987, PASP, 99, 116Green, J. D., Hartmann, L., Calvet, N., et al. 2006, ApJ, 648, 1099Hartmann, L. & Kenyon, S. J. 1996, ARA&A, 34, 207Herbig, G. H. 1977, ApJ, 217, 693Hogerheijde, M. R., van Dishoeck, E. F., Blake, G. A., & van Langevelde, H. J.1998, ApJ, 502, 315Hogerheijde, M. R., van Dishoeck, E. F., Salverda, J. M., & Blake, G. A. 1999,ApJ, 513, 350Jørgensen, J. K., Hogerheijde, M. R., van Dishoeck, E. F., Blake, G. A., &Sch¨oier, F. L. 2004, A&A, 413, 993Kenyon, S. J. & Hartmann, L. 1989, ApJ, 342, 1134Langer, W. D. & Penzias, A. A. 1993, ApJ, 408, 539Levreault, R. M. 1983, ApJ, 265, 855Levreault, R. M. 1988, ApJS, 67, 283Matthews, B. C., Hogerheijde, M. R., Jørgensen, J. K., & Bergin, E. A. 2006,ApJ, 652, 1374McMuldroch, S., Sargent, A. I., & Blake, G. A. 1993, AJ, 106, 2477Momose, M., Ohashi, N., Kawabe, R., Nakano, T., & Hayashi, M. 1998, ApJ,504, 314Ossenkopf, V. & Henning, T. 1994, A&A, 291, 943Quanz, S. P., Henning, T., Bouwman, J., et al. 2007, ApJ, 668, 359Richardson, K. J., White, G. J., Avery, L. W., Lesurf, J. C. G., & Harten, R. H.1985, ApJ, 290, 637Sandell, G. & Weintraub, D. A. 2001, ApJS, 134, 115Scoville, N. Z., Sargent, A. I., Sanders, D. B., et al. 1986, ApJ, 303, 416Turner, N. J. J., Bodenheimer, P., & Bell, K. R. 1997, ApJ, 480, 754van Kempen, T. A., van Dishoeck, E. F., Salter, D. M., et al. 2009, A&A, 498,167Vorobyov, E. I. & Basu, S. 2006, ApJ, 650, 956Weintraub, D. A., Sandell, G., & Duncan, W. D. 1991, ApJ, 382, 270