Trigonometric Distance and Proper Motion of IRAS 20056+3350: Massive Star Forming Region on the Solar Circle
Ross A. Burns, Takumi Nagayama, Toshihiro Handa, Toshihiro Omodaka, Akiharu Nakagawa, Hiroyuki Nakanishi, Masahiko Hayashi, Makoto Shizugami
aa r X i v : . [ a s t r o - ph . S R ] O c t Accepted for publication in:
The Astrophysical Journal.
Trigonometric Distance and Proper Motion of IRAS 20056+3350:Massive Star Forming Region on the Solar Circle
Ross A.
Burns , Takumi
Nagayama , Toshihiro
Handa , Toshihiro
Omodaka , Akiharu
Nakagawa , Hiroyuki
Nakanishi , Masahiko
Hayashi , Makoto
Shizugami , Graduate School of Science and Engineering, Kagoshima University,1-21-35 Kˆorimoto, Kagoshima, Kagoshima 890-0065, [email protected] Mizusawa VLBI Observatory, National Astronomical Observatory of Japan,2-21-1 Osawa, Mitaka, Tokyo 181-8588, Japan National Astronomical Observatory of Japan,2-21-1 Osawa, Mitaka, Tokyo 181-8588, Japan School of Mathematical and Physical Science, The Graduate University for Advanced Studies(SOKENDAI), Hayama, Kanagawa 240-0193, Japan Mizusawa VLBI Observatory, National Astronomical Observatory of Japan,2-12 Hoshi-ga-oka, Mizusawa-ku, Oshu, Iwate 023-0861, Japan (Received ; accepted )
Abstract
We report our measurement of the trigonometric distance and proper motion of IRAS 20056+3350,obtained from the annual parallax of H O masers. Our distance of D = 4 . +0 . − . kpc, which is 2.8 timeslarger than the near kinematic distance adopted in the literature, places IRAS 20056+3350 at the leadingtip of the Local arm and proximal to the Solar circle. Using our distance we re-evaluate past observationsto reveal IRAS 20056+3350 as a site of massive star formation at a young stage of evolution. This result isconsistent with the spectral energy distribution of the source evaluated with published photometric datafrom UKIDSS, WISE, AKARI, IRAS and sub-millimetre continuum. Both analytical approaches revealthe luminosity of the region to be 2 . × L ⊙ , and suggest that IRAS 20056+3350 is forming an embeddedstar of ≥ ⊙ . We estimated the proper motion of IRAS 20056+3350 to be ( µ α cos δ , µ δ ) = ( − . ± . − . ± .
52) mas yr − from the group motion of H O masers, and use our results to estimate the angularvelocity of Galactic rotation at the Galactocentric distance of the Sun, Ω = 29 . ± .
29 km s − kpc − ,which is consistent with the values obtained for other tangent point and Solar circle objects. Key words:
Masers - Stars; individual (IRAS 20056+3350) - Galaxy: structure and dynamics
1. Introduction
The morphology of a galaxy is determined visually andis highly influenced by the distribution of massive starswhich are bright and typically confined to the spiral arms(Urquhart et al., 2014; Reid et al., 2014). Although theexclusivity of massive star formation to the spiral arms isnot yet understood it does bring to light a strong inter-play between the spiral pattern itself and the conditionsrequired to trigger massive star formation. Thus, neitherphenomena can be understood independently; we cannotexplain massive star formation without explaining the roleof the spiral pattern and we require the by-products ofmassive star formation to understand the morphology ofgalaxies by tracing the arms.From our viewpoint within the Galactic disk, perhapsthe most important parameter common to both fields ofinvestigation is distance. In addition to the clear necessityof accurate distances for mapping the Galactic arms, theinterpretation of the physical properties of a massive starforming region (MSFR) is greatly influenced by the dis-tance estimate adopted - since many of these parameters have a non-linear relationship with distance.For convenience, many investigators make use of thekinematic distances of Galactic star forming regions(SFRs). However, kinematic distance calculations requirepre-calibration of the Galactic rotation curve and sufferfrom distance ambiguities ( D near and D far ) for objects inthe inner Galaxy. Moreover, they are unreliable when thesource has a large peculiar velocity with respect to purelycircular rotation in the Galactic disk. Consequently, thekinematic distance can be poorly determined for objectsin some particular locations such as in the direction of theGalactic center and anticenter, and Local Arm.Measurements of trigonometric parallaxes to GalacticMSFRs directly improves the reliability of their distances.Such experiments involve measurement of the annual par-allax of MSFRs typically via astrometric monitoring ofmaser spots using VLBI networks such as the VLBIExploration of Radio Astrometry (VERA), the Very LongBaseline Array (VLBA), and the European VLBI Network(EVN). In addition to distance measurement, masers arealso extremely useful for understanding the processes ofstar formation itself since masers emitted by different R. A. Burns et al. [Vol. ,molecular species and transitions are known to be asso-ciated with different physical environments (Bartkiewicz& van Langevelde, 2012). As a result, masers are of-ten seen to trace structures such as expanding shells andshock fronts (Trinidad et al., 2013), and bipolar outflows(Imai et al. 2007; Nagayama et al. 2008; Moscadelli et al.2011; Torrelles et al. 2014 and G236.81+1.98 in Choi et al.2014). Masers thus allow the internal motions of star for-mation mechanisms to be seen even though the structuresthemselves may not be directly unobservable due to theembedded nature of such systems.Finally, combining the proper motions and line of sightmotions of masers sometimes allows estimation of thethree-dimensional secular motions of a star forming regionin the plane of the Galaxy. This is one of the unique ad-vantages of VLBI maser investigations and is a dominantapproach to understanding the kinematics and structureof the Milky Way Galaxy (MWG) via the evaluation of theGalactic constants, R , Θ and Ω (Honma et al., 2012;Reid et al., 2014). These parameters can be evaluatedmore reliably for SFRs that reside at special locations inthe MWG such as the tangent points (Nagayama et al.,2011; Burns et al., 2014) and Solar circle (Ando et al.,2011).All VLBI observations discussed in this paper were car-ried out using VERA (Kobayashi et al., 2003) which is aJapanese VLBI array dedicated to measuring the annualparallax of maser sources in the MWG. IRAS 20056+3350was chosen for this investigation for its bright and stablemaser emission at 22 GHz - first seen in Jenness et al.(1995) - and its presumed proximity to the Solar circle.IRAS 20056+3350 is listed in the IRAS catalogue of pointsources (Beichman et al., 1988).In this work we aim to make a contribution to eachof the aforementioned topics. This paper continues asfollows: Observations and data reduction are discussed in §
2. Results are reported in §
3, including analyses of theastrometric accuracy achieved in parallax fitting of maserspots. The physical nature of the IRAS 20056+3350MSFR is explored in § using the results obtained fromour observations with VERA. Conclusions are reportedin §
2. Observations and data reduction
Data were obtained using VERA in dual-beam mode.By observing H O masers in IRAS 20056+3350 andthe J2010+3322 reference continuum source simultane-ously we calibrated tropospheric phase fluctuations us-ing the reference source and applied solutions directlyto the maser data in real-time without interpolation.The scan integration time of the pair was about 9 min-utes. Intermittent observations of BL Lac or 3C454.3were made every 1.5 hours for bandpass calibration. A typical observation session lasted roughly 8 hours, pro-viding ∼ α, δ ) J2000 . =(20 h m s . ◦ α, δ ) J2000 . =(20 h m s . ◦ seeSection 3.1 ). Table 1.
Summary of observations made with VERA.
Observation DetectedEpoch Date spots1 2012 Feb 05 22 2012 Apr 30 23 2012 Aug 09 24 2013 Feb 16 35 2013 Apr 22 36 2013 May 05 27 2013 Dec 24 2
Interferrometric correlation was carried out using theMitaka FX correlator (Chikada et al., 1991). A frequencyresolution of 15.625 kHz, corresponding to a velocity res-olution of 0.21 km s − , was used for the maser data, andfor the continuum source a frequency resolution of 250kHz was used. We did not use the central IF of the con-tinuum source data during data reduction since the fre-quency resolution of this IF channel did not match that ofthe other 14 IFs. This is an affect of the correlation pro-cess whereby the frequency resolution of this IF reflectsthat of the maser data. The small loss of bandwidth doesnot effect our goals adversely since the continuum sourceis readily detected with a signal to noise ratio of ≥ Table 2.
The general properties of H O maser in IRAS 20056+3350 detected with VERA.
Spot V LSR
Detected ∆ α cos δ ∆ δ π µ α cos δ µ δ ID (km s − ) epochs (mas) (mas) (mas) (mas yr − ) (mas yr − )A +7 .
27 1234567 0 0 0 . ± . − . ± . − . ± .
03B +7 .
27 12345** − .
03 +0 .
65 0 . ± . − . ± . − . ± . − .
25 ****456* − .
69 +581 . − . ± . − . ± .
03D +0 .
25 *******7 +804.61 +3.47Group fitting 0 . ± . − . ± . − . ± . ments from the Japan meteorological association (JMA)to the same effect, though at a lower time resolution.The reliability of these procedures is explored in detailin Nagayama (2014, in preparation).All data were reduced using Astronomical ImageProcessing System (AIPS) developed by the NationalRadio Astronomy Observatory (NRAO). Amplitude andbandpass calibration for both beams was carried out usingthe standard procedures of reduction of astrometric VLBIdata in AIPS. During the first round of data reductionwe created phase referenced images of the maser emissionby calibrating maser data using phase solutions obtainedfor J2010+3322 in the AIPS task FRING. The positionsof masers were determined by the peaks of 2D Gaussianfits applied to the final images. Using this procedure wefound the two brightest maser spots (spot ID:A and spotID:B; see Table 2 ). The other spots (spot ID:C and spotID:D) were only seen in self-calibrated images created inthe second round of data reduction. In this reduction pro-cedure we made single-beam images by self-calibration us-ing the AIPS task pair IMAGR and CALIB on the bright-est maser spot (spot ID:A). Subsequent maser positionswere determined relative to the astrometric position ofthe bright reference maser, which was determined fromthe phase-referenced map.
3. Results O masers in IRAS 20056+3350
A modest total of 4 individual maser spots were de-tected in our observations. Maser spot detections aresummarised in Table 2 where position offsets are givenrelative to the reference maser (Spot ID:A) for which wemeasure first epoch absolute co-ordinates of ( α,δ ) J2000 . =(20 h m s . ◦ − .
25 to +7 .
27 km s − . Thus, all detected masersare blueshifted with respect to the parent cloud - whose ve-locity from molecular line observations is known to be +9km s − ( see section 4.2.1 ). The spatial distribution andvelocities of maser spots detected in IRAS 20056+3350 arepresented in Fig. 4 with arrows indicating proper motionvectors ( see section 3.4 ). Parallax and proper motion fitting was performed si-multaneously on the two maser spots which were identi- fiable in multiple observations, spanning more than oneyear. Astrometric motions of these spots were decon-structed into linear and sinusoidal components arisingfrom the sky-plane proper motion and annual parallax,respectively. In the fitting procedure, nominal fitting er-rors in the R.A. and Dec. directions were applied andreduced itteratively until a χ value of unity was reached.These error floors were 0.227 mas in R.A. and 0.050 masin Dec. -12-10-8-6-4-2 0 2 -6-4-2 0 2 D e c l . o ff s e t ( m a s ) R.A. offset (mas) -16-14-12-10-8-6-4-2 0 2012 2013 2014 O ff s e t ( m a s ) Year -1.2-1-0.8-0.6-0.4-0.2 0 0.2 0.4 0.6 2012 2013 2014 O ff s e t ( m a s ) Year
Fig. 1.
Parallax and proper motion fitting for two maserspots in IRAS 20056+3350. Spot ID:A and ID:B are repre-sented by black and grey points respectively. Left: Sky-planemotion of the maser spots. Middle: motion in R.A. (above)and Dec. (below) as a function of time. Right: parallac-tic motion in R.A. (above) and Dec. (below) after subtrac-tion of the linear proper motion (result for the combined fit).
When data were fit using the two most stable andbrightest maser spots (Spot ID:A and ID:B) togetherwe derived an annual parallax of π = 0 . ± .
026 mas( uncertainty of . D = 4 . +0 . − . kpc. Results of the fitting procedure are il-lustrated in Fig. 1. When repeating the fitting procedurefor these spots individually we arrived at consistent valuesfor both the distance and error floors ( see Table 2 ). It isclearly seen in Fig. 1 that the data were fit well in the Dec.direction and poorly in the R.A. direction. The cause ofthis situation in the context of astrometric accuracy isrevisited in the next subsection.Molinari et al. (1996) were the first to estimate the kine-matic distance to IRAS 20056+3350. Assuming that thesource is at the near kinematic distance for Galactic ro-tation, and using the IAU recommended Galactic con-stants of R = 8 . = 220 km s − , they derived D = 1 .
67 kpc from the radial velocity of of their NH lineobservations at v LSR = 9 . − . To date, this value R. A. Burns et al. [Vol. ,has been adopted in all subsequent literature regardingthis source. Assuming Galactic constants R = 8.05 kpcand Θ = 235 km s − (at V ⊙ = 15.3 km s − ; Honmaet al. 2012) and the condition of flat Galactic rotation Θ= Θ = 235 km s − , we calculate the near and far kine-matic distances of IRAS 20056+3350 as D near = 1.2 kpcand D far = 3.9 kpc respectively. Our distance obtainedvia trigonometric parallax is close to the far kinematicdistance. M illi a r c s e c o nd s Milliarcseconds epoch 1 epoch 2epoch 3epoch 4epoch 5epoch 7
Fig. 2.
Multi-epoch phase-referenced images of the promi-nent maser spots (
Left
Spot ID:A and
Right
ID:B). Thebackground emission from noise has been removed artificiallyto highlight the emission of maser spots used in our anal-ysis. Maser spots move from center to south-west, chrono-logically (epoch numbers 1 to 7 are shown, however epoch 6was omitted for its proximity to epoch 5 in time, which con-fused the image - Spot ID:B was not detected in epoch 6).Contours shown are multiples of 3, 5, 10, 20 and 30 of theroot mean squared noise for individual maps. The origin ofthe map is ( α,δ ) J2000 . = (20 h m s . ◦ α, δ ) J2000 . = (20 h m s . ◦ In the parallax fitting results of Fig. 1, the astrometricpositions in the R.A. direction indicate a notably largedegree of dispersion. Upon inspection of the brightnesspeaks in the emission maps, which we present in Fig. 2,emitting regions exhibit structural elongation primarilyin the R.A. direction. To investigate the astrometric errorcontribution from the structure of masers we monitoredthe spatial separation between the two bright, persistentmaser spots. In the absence of acceleration, relative mo-tions should be linear, thus we can compare the astromet-ric accuracy in R.A. and Dec. by evaluating the deviationfrom linear best fit motions in each respective direction. S p o t s e p a r a t i o n ( m a s ) Time (Days)
Fig. 3.
Spatial separation of the two prominent maser spots(Spot ID:A and ID:B) in R.A. (above) and Dec. (below)over the first five epochs in which both maser spots were de-tected. In the absence of acceleration, the relative motionsof maser spots should manifest as a linear increase or de-crease in spot separation over time. Best fits to the spotseparation in R.A. and Dec. are overlain. Standard de-viations to the best fits in R.A. and Dec. directions were0.19 mas and 0.04 mas respectively. The first side-loberadius, as discussed in the text, appears near 2.5 mas inour data, and thus only affects the R.A. separation.
In Fig. 3, the large deviations from linearity seen in R.A.reveal an instability in the astrometric accuracy. Thesedeviations are likely caused by the notable elongation ofmaser structures, which results in an inaccurate determi-nation of emission peaks in the interferrometric images.The perceived elongation may be real structure, or it maybe an apparent effect caused by the smearing of the twospots at small separation from each other. Furthermore,the spatial separation of maser spots was often close tothe first side-lobe radius in the R.A. direction (about 2.5mas in our VERA observations). Interaction of the maseremission with side-lobes is likely to have contributed ad-ditional smearing of emission peaks in the R.A. direction.The standard deviation values for the linear fits in R.A.and Dec. were 0.19 and 0.04 respectively. These valuesare similar to the error floors arrived at in the parallax fit-ting procedure which were required to reach a χ value ofunity. Thus, our analysis suggests that the maser struc-ture contributes a dominant source of astrometric errorin these observations. Although this effect is detrimentalfor fitting the maser parallax in the R.A. direction, theastrometry in the Dec. direction is altogether unaffected,i.e this error is confined to the R.A. direction and thussupports preferential use of the Dec. direction offsets asthe most suitable approach to parallax fitting. O masers
We found a third maser spot (spot ID:C) in 3 consec-utive epochs of the self-calibrated images. In order todeconvolve parallactic and secular motions a maser via fit-ting, a spot must be detected for around 1 year or longerin the phase-referenced images. However, we were able too. ] Distance to IRAS 20056+3350: MSFR on the Solar circle 5estimate the proper motion of this spot via its relative mo-tion to the bright maser spot (spot ID:A) detected in theself-calibrated images by assuming that both spots havea common annual parallax. By combining the relativemaser motions with the known motion of the referencemaser we obtained the sky-plane proper motion of thethird spot, where the error is calculated as the quadra-ture sum of the standard deviation in the relative motionof masers, and the proper motion error of the referencemaser from parallax fitting. A fourth maser spot (spotID:D) was detected in the final observation epoch onlyand thus its motion could not be found. Maser propermotions and their errors are listed in Table 2. Absolutemaser proper motions are shown in Fig. 4 where arrowsrepresent velocity vectors. -1000-500 0 500 1000 D e c l . o ff s e t ( m a s ) R.A. offset (mas) -2-1 0 +1 +2 +3 +4+5+6 +7 L S R v e l o c i t y ( k m / s ) -500 0 500 1000 6 mas/yr Fig. 4.
Distribution and absolute sky-plane propermotions of H O maser spots in IRAS 20056+3350.At D = 4 .
69 kpc typical motions of ∼ − correspond to velocities of ∼
135 km s − . No obvious spatial nor velocity structure can be inferredfrom our maser vector maps alone. With the small numberof maser spots it is unwise to search our images for self-contained geometric structures arising from phenomenasuch as bipolar outflows and spherical expansions; suchanalysis is powerful only if the number of maser spots islarge enough to reveal structure in the velocity field inregions of star formation ( for example
Imai et al. 2011).
4. Discussion
Zhang et al. (2005) presented an on-axis outflowmapped in CO which had a velocity centered at about+9 km s − , in agreement with other molecular line obser-vations ( see Table 3 ). The spectrum exhibits a triple-peakmorphology with emission detected in the range of 0 ∼ +20km s − . The peaks correspond to the blue-shifted outflow,the parent cloud and redshifted outflow, in ascending ve-locity. Since the V LSR of our masers are all consistent withthe blue limit of the CO emission we conclude that ourobservations are likely sampling masers associated withthe blueshifted lobe of the outflow, which appears alignedto the line of sight. Although this does not allow directinterpolation of maser motions to a kinematic center, as is demonstrated in Imai et al. (2011), it does allow us tomake the reasonable assumption that the sky-plane propermotions of maser spots with respect to the driving sourceare small, since the largest velocity components can beexpected along the line of sight, in agreement with thedirection of the molecular outflow. As such, the averageproper motion of maser spots should give a reasonableapproximation to the secular motion of the SFR.We evaluated the group averaged proper motion for allmaser spots in IRAS 20056+3350 to be ( µ α cos δ , µ δ ) =( − . ± . − . ± .
52) mas yr − . Error values arethe standard error of the mean, σ / √
3, where σ is thestandard deviation of the proper motion of 3 spots. Aftersubtraction of the average group motion, residual motionsreveal the internal kinematics of H O masers in IRAS20056+3350. We present a map of the internal masermotions in Fig. 5.The proper motions of SFRs can be used to break thedistance ambiguity associated with kinematic distance es-timates because sources at near and far kinematic dis-tances are expected to exhibit different proper motions onthe sky. For IRAS 20056+3350 the near and far kinematicdistances are D near = 1.2 kpc and D far = 3.9 kpc, respec-tively. Sources at these locations should show the propermotions of µ l = − . − and µ l = − . − respectively, if they rotate circularly around the Galacticcenter with Θ = 220 km s − and assuming R = 8 .
05 kpc(Honma et al., 2012). -500 0 500 1000 -1000-500 0 500 1000 D e c l . o ff s e t ( m a s ) R.A. offset (mas)1 mas/yr -11 -9 -7 -5 -3 -1+1 V e l o c i t y d i ff e r e n c e ( k m / s ) Fig. 5.
Internal motions of H O maser spots in IRAS20056+3350. The systemic motion is assumed to be theaverage proper motion of all measured maser spots onthe sky, and the radial velocity of the ambient molec-ular gas, 9 km s − . At D = 4 .
69 kpc typical mo-tions of ∼ − correspond to ∼
22 km s − . Using our estimate of the systemic proper motion ofIRAS 20056+3350, ( µ α cos δ , µ δ ) = ( − . ± . − . ± .
52) mas yr − , we calculated the sky-plane motion of thesource with respect to the Solar LSR in the direction ofGalactic longitude as µ l cos b = − . ± .
48 mas yr − ,using the standard solar motion (U ⊙ ,V ⊙ ,W ⊙ ) = (+10.3,+15.3, +7.7) km s − (Kerr & Lynden-Bell 1986, see also Ando et al. 2011). This proper motion value is consistentwith that estimated for a source at the far kinematic dis-tance. Furthermore, the far kinematic distance, D far = R. A. Burns et al. [Vol. , Table 3.
Systemic velocity, velocity width, and beam size of IRAS 20056+3350 observations.
Transition v LSR ∆ v Beam size
Ref ( kms − ) ( kms − ) (arcsec)CS J = (2 −
1) +8.8 3.2 39 Bronfman et al. (1996)CS J = (5 −
4) +9.0 3.2 21 Jenness et al. (1995)C O J = (2 −
1) +9.0 3.0 21 Jenness et al. (1995) CO J = (1 −
0) +8.56 4.9 55 Wu et al. (2001)NH (1 ,
1) +9.4 2.29 40 Molinari et al. (1996)NH (2 ,
2) +9.0 1.82 40 Molinari et al. (1996) D tri = 4 .
96 kpc, mea-sured using annual parallax compared to the near distance D near = 1 . Since Molinari et al. (1996), all astrophysical works re-garding IRAS 20056+3350 adopt the near kinematic dis-tance of 1.67 kpc in analysis of their data. Our trigono-metric distance of D = 4 . +0 . − . kpc is 2.8 times larger,thus significantly impacting the interpretation of data col-lected up to now. As such, we briefly revisit these past ob-servations to summarise the nature of IRAS 20056+3350 -re-evaluated using the trigonometric distance and includealso relevant distance-independent results to provide a fullaccount of this MSFR. Wood & Churchwell (1989) developed a diagnostic toolused to identify embedded massive OB type stars usingIRAS point source catalogue colour criteria. Using theirmethod IRAS 20056+3350 is a candidate for harbouringat least one embedded massive star.The infrared luminosity of IRAS 20056+3350 fromthe total IRAS fluxes was estimated as L
IRAS = 1100L ⊙ (Casoli et al., 1986) using an assumed distance of 1.0kpc. At the new distance, 4.69 kpc, it is revised asL IRAS = 24000 L ⊙ . The bolometric luminosity of IRAS20056+3350 should be close to L IRAS , assuming that mostof the emission from the embedded central star is ab-sorbed and then re-emitted at infrared wavelengths. Using M ∝ L . (Allen, 1976), as was done in Casoli et al. (1986),the mass of a central star is revised from 7 . M ⊙ to 16.5M ⊙ , although the authors did not present the mass ofthe star. Casoli et al. (1986) also produced a map in CO J (1 −
0) with 4.’4 resolution using the Bordequx 2.5m telescope and estimated a gas mass of M H2 = 56 M ⊙ assuming local thermodynamic equilibrium (LTE), a dis-tance of D = 1 kpc and an abundance ratio of CO/H of2 × − from Dickman (1978). The molecular gas massis revised from 56 M ⊙ to M H2 = 1200 M ⊙ using the sameassumption as that of Casoli et al. (1986) but the reviseddistance.Far infrared (FIR) continuum maps of IRAS20056+3350 at 450 µ m and 800 µ m from Jennesset al. (1995) clearly show dense cores, indicative ofembedded star formation. Their Figures 1 and 2 showthat the far infrared emission is associated with both the IRAS point source and water maser source, although theemission peak at 450 µm is offset from the masers by 5”in the direction of the cluster center. Although radiocontinuum emission at 8 GHz was marginally detected intheir VLA observation, no compact source was identified(Figure 6 in Jenness et al. 1995). This suggests that noH II region is present, although Jenness et al. (1995) donot comment explicitly on this.To establish the radial velocity of the system weused published thermal molecular line observations sincethese typically trace ambient molecular gas. The sys-temic velocities, velocity widths, and telescope beamsizes from single-dish molecular line observations of IRAS20056+3350 are summarised in Table. 3. From these re-sults we nominally conclude the LSR velocity of the sys-tem to be +9 . ± .
25 km s − , where the error is thestandard deviation of the listed values.NH (1,1) and (2,2) line observations of IRAS20056+3350 were carried out by Molinari et al. (1996).The measured line widths were ∆ v (1 , = 2 .
29 km s − and∆ v (2 , = 1 .
83 km s − , respectively. The NH (1 ,
1) and(2 ,
2) lines predominantly trace the less dense envelopeand the dense core, respectively. The authors argue thatsuch sources, that exhibit lower velocity dispersion in thecentral core compared to the outer envelope, are charac-teristic of the lack of an ultra compact H II (UCHII) region,and possibly precede this stage.Zhang et al. (2005) mapped IRAS 20056+3350 in CO J = (2 −
1) using the NRAO 12m telescope and reporteda molecular outflow centered at the position of the IRASsource. Lobes of blueshifted and redshifted emission areconcentric in the sky-plane indicating that the outflow isorientated to a pole-on geometry. Three velocity compo-nents are easily distinguishable in their spectra (panel ID115 in Fig.2 of Zhang et al. 2005). These componentsare symmetric about the center in the line profile: theblueshifted component which peaks at +5 km s − , theparent core at +9 km s − , and the redshifted componentwhich peaks at +13 km s − . Emission was detected overthe range of 0 to +20 km s − . From their CO J = (2 − outflow = 2 . ⊙ , p outflow = 20 . ⊙ km s − andE outflow = 2 . × ergs. Our new distance gives outflowparameters M outflow = 18 M ⊙ , p outflow = 163 M ⊙ km s − and E outflow = 2 . × ergs.o. ] Distance to IRAS 20056+3350: MSFR on the Solar circle 7 Table 4.
Flux parameters used in the SED fitting of IRAS 20056+3350.
Telescope Band Photometry Aperture size Source name Mission reference(name or µ m) (magnitude or flux) (arcsec)UKIDSS J 18 . ± .
105 mag 3 J200731.43+335937.6 Lucas et al. (2008)H 15 . ± .
012 mag 3 J200731.38+335939.4K s . ± .
002 mag 3WISE 3.4 7 . ± .
024 mag 8.5 J200731.37+335940.9 Cutri et al. (2014)4.6 4 . ± .
04 mag 8.512 3 . ± .
016 mag 8.522 − . ± .
013 mag 16.5AKARI 9 5510 ± . ±
211 mJy 6IRAS 60 422 ± . ± .
12 Jy 120JCMT 450 33 ± . . ± . A practical method for investigating the physical na-ture and evolutionary stage of an astronomical object isto draw information from its spectral energy distribution(SED). In this effort we constructed the SED of IRAS20056+3350 using published data.Infrared fluxes were extracted from point source cata-logues using the NASA Infrared Science Archive (IRSA),from the telescope missions of UKIDSS, WISE, AKARI,and IRAS. We also use JCMT fluxes at 450 and 800 µ mfrom Jenness et al. (1995).The aperture sizes or angular resolutions of each of thecatalogs were different. In Figure. 6 we show a three-colour image of IRAS 20056+3350 made using publisheddata from UKIDSS in J, H, K s bands, and whose cat-alogue has the the highest angular resolution of any weused. A distinctly redder region is seen in the center-western portion of the cluster. A star forming core withredder colours has more intense radiation at the longerwavelengths which is characteristic of cores which areforming more massive stars. We identify this region as themost likely site of embedded massive stars. Furthermore,a dark lane which is prominent in the K -band (Varricattet al., 2010) and the association of H O masers supportthat this is a site of embedded star formation. Thus, we re-quired that flux apertures from all point source cataloguesbe positionally consistent with this region. Details of thefluxes used in SED fitting are given in Table 4 and theposition and sizes of apertures are overlaid onto Figure. 6.Although the beams of the larger apertures will invari-ably contain multiple cluster members, we made no cor-rections for this because our target should be the brightestin the beam; our target is the most massive and the fluxfrom other objects should be negligible, since the lumi-nosity is extremely sensitive to the mass of the object, bythe relation L ∝ M . .To test the reliability of using the IRAS 60 and 100 µ m bands in our SED we compared the photometries ofIRAS F µ m and WISE F µ m and found them to be con-sistent, supporting the reliability of using the IRAS fluxes.450 and 800 µ m fluxes, measured using the James Clarke . : : . . . . : : . Right ascension D e c l i n a t i o n ALLWISE W4ALLWISE W1,2,3AKARI IRC UKIDSS
Fig. 6.
Aperture size and positions taken from point sourcecatalogues of various infrared telescope missions. These cor-respond to photometric measurements used to evaluate theSED of IRAS 20056+3350, as is discussed in the main text.Apertures are plotted over the near-IR composite image fromUKIDSS where colours blue, green and red correspond tobands J, H and K s . Proper motions of H O masers ob-served with VERA are indicated by white arrow vectors. TheIRAS aperture was too large to be displayed in this figure.
Maxwell telescope (JCMT) by Jenness et al. (1995), wereevaluated with an aperture equivalent to the extent of theFWHM of the observed emission. Inclusion of these fluxesmay lead to an overestimation of the integrated SED fluxsince hot gas associated with the outflow will be included.However, since the core emission is compact and centeredon the maser source, the flux is likely dominated by theembedded massive star. These data provide valuable con-fines for the longer wavelength portion of the SED. R. A. Burns et al. [Vol. ,To interpret our photometric data we use the fittingtool of Robitaille et al. (2007) which is well documentedin the introductory paper referenced here. The SED ofIRAS20056+3350, according to the model fitting is shownin Fig. 7.The SED model corresponds to a 18.4 M ⊙ central starof age 10 yrs, embedded in an envelope of mass 3300M ⊙ . The total luminosity of the region is 2 . × L ⊙ .The luminosity and stellar mass derived from the SEDmodel agrees well with the estimates made using thefour IRAS bands by of Casoli et al. (1986) re-evaluatedusing our distance estimate as L IRAS = 24000 L ⊙ and M ∗ ∼ . M ⊙ . Furthermore, the envelope mass derivedfrom the SED model agrees within a factor of two withthe estimate of the molecular hydrogen content made byCasoli et al. (1986) re-evaluated using our distance esti-mate as M H = 1200 M ⊙ . Fig. 7.
SED of IRAS 20056+3350 using datafrom UKIDSS, AKARI, WISE, IRAS and JCMT.
Interestingly, the inclination angle required to producethe best-fit model is 18 ◦ to the line of sight, which agreeswith the general orientation inferred from the CO J =(2 −
1) observations of Zhang et al. (2005). The ability ofthe SED fitting software to determine the inclination ofyoung stellar object (YSO) systems solely from photom-etry data has been demonstrated for G34.4+0.23MM inRobitaille (2008) and for IRAS 04368+2557 in Robitailleet al. (2007). In these cases the system inclination hadbeen previously established observationally from the ori-entation of bipolar outflows and was well matched by theSED fitting tool.It should be pointed out that the successful fit of theSED model to the photometry data shows only that thereexists a theoretical model consistent with the data and itis not proof that the model results truly reflect the natureof the source. The parameters directly determined fromthe data are luminosity and temperature of the emittingregion, the further details (stellar mass, envelope mass,inclination, etc.) depend on the evolutionary tracks ap-plied to the data, this is cautioned in Robitaille (2008). However, in our case the SED results are consistent withthe picture drawn from an abundance of data from pastobservations re-evaluated at the trigonometric distance.It is this consistency rather than the pleasing appearanceof the model fit to the data, that raises confidence in ourresults. These consistencies, along with the common pro-posed inclination angle, supports the basis of our argu-ment; that the secular proper motion of IRAS 20056+3350can be reasonably estimated from the group motions ofH O masers associated with the line of sight outflow.We conclude that IRAS 20056+3350 is a distant site ofmassive star formation at a young stage in evolution. Theexistence of an embedded, young massive star is stronglysuggested by the IRAS colours, SED, and presence a com-pact sub-millimeter core. With respect to youth, the evo-lutionary stage precedes the formation of an ultra compactH II (UCHII) region, this is suggested from the turbulenceratio of the core and envelope seen in NH and is corrob-orated by the lack of significant centimeter structure at8GHz. Line spectra of core tracing molecular transitionssuch as CS reveal substantial turbulence ( see Table. 3)which is suggested by Ao et al. (2004) to be characteristicof young SFRs which have not yet dissipated primordialturbulent motions inherent in the core. An active bipo-lar molecular outflow, seen in spectral wings and emissionmap, appears aligned to the line of sight - which we believeto be associated with the H O masers we observed withVERA due to the consistency of velocities. The overallpicture ascertained from past observations is corroboratedby the SED of IRAS20056+3350, which resembles a YSOin the Class I evolutionary phase.
Full-scale studies of the structures of the spiral armshave recently been made available from analysis of thejoint results of VLBI annual parallax measurements ofMSFRs. Recent cases are; the Local Arm by Xu et al.(2013), the Perseus Arm by (Zhang et al., 2013; Choiet al., 2014), the Saggitarius Arm by Wu et al. (2014) andthe general structure is discussed in Reid et al. (2014).IRAS 20056+3350 is the most distant MSFR in the Localarm for which a trigonometric distance has been deter-mined. Therefore, although we cannot justify a large-scalere-evaluation of the Local arm, we can provide an impor-tant contribution to the picture recently provided by theaforementioned authors. The nature of the Local armis discussed particularly in the work of Xu et al. (2013)where the authors consider the three possible scenarios; That the Local Arm is a branch of the Perseus Arm. That the Local Arm may be part of a major arm andconnects to the Carina Arm. That it is an independentarm segment - a ‘spur’.In Fig. 8 we show the position of IRAS 20056+3350in the context of the Galactic spiral structure, which wasrecently compiled by Reid et al. (2014) (
References forindividual SFRs therein ). In drawing logarithmic spiralarms they used the Galactocentric distance of the Sun,R = 8 .
34 kpc.o. ] Distance to IRAS 20056+3350: MSFR on the Solar circle 9
IRAS 20056+3350
Fig. 8.
Location of IRAS 20056+3350 in relationto the Galactic spiral structure as evaluated us-ing VLBI observations of MSFRs, compiled by Reidet al. (2014) ( individual source references therein ). As seen in Fig. 8, the Galactic location of IRAS20056+3350 diverges somewhat from the logarithmiccurve determined for the Local arm by Reid et al. (2014).Deviation is in the direction of the Perseus arm and followsthe example of three other MSFRs at the leading tip ofthe arm; G075.76+00.33 (Xu et al., 2013), G075.78+00.34(Ando et al., 2011; Xu et al., 2013) and AFGL 2591 (Ryglet al., 2012). The location of IRAS 20056+3350 is gener-ally consistent with what would be expected in the sce-nario where the Local Arm joins onto the Perseus armfurther down into the spiral pattern - scenario of Xuet al. (2013). However, the alternative scenarios cannotbe ruled out until this region of the Galaxy is mapped inmore detail. Ω Galactic constant
From our trigonometric distance we calculated theGalactocentric distance of IRAS 20056+3350 to be R ∗ =7 .
91 kpc, when using R = 8.05 kpc from Honma et al.(2012). Since R ∗ ≃ R , and from the low value of v LSR = 9km s − , it is evident that IRAS 20056+3350 resides spa-tially and kinematically near to the Solar circle.Nagayama et al. (2011) demonstrated that the angu-lar velocity of Galactic rotation at the Galactocentric dis-tance of the Sun, Ω , can be estimated for objects near theSolar circle and tangent points, with relaxed assumptionson the adopted value of R . For objects with negligible peculiar motion, the angular velocity of Galactic rotationat the Sun is given byΩ = − a µ l + v r (cid:18) D tan l − R sin l (cid:19) (1)where D is the distance to the source from the Sun, v r isthe LSR velocity, µ l is the systemic sky-plane proper mo-tion in the direction of Galactic longitude, and a = 4 . − kpc − (mas yr − ) − is a unit conversion factor.In order to make use of Equation (1) we first convertour systematic proper motion estimate into the propermotion with respect to the LSR in the Galactic coordinatesystem. To correct for the motion of the Sun with respectto the LSR we use the following values: (U ⊙ ,V ⊙ ,W ⊙ ) =(+10.3, +15.3, +7.7) km s − (Kerr & Lynden-Bell 1986, see also Ando et al. 2011). The relevant parameters,determined in this paper, for IRAS 20056+3350 are: D = 4 . +0 . − . kpc and µ l = − . ± .
48 mas yr − . Weuse v r = +9 . ± .
25 km s − determined from molecularline observations ( see section 4.2.1 ). Applying the aboveparameters to Equation (1), and using R = 8.05 kpc,givesΩ = 29 . ± .
29 km s − kpc − We demonstrate the invariability of this approach on theadopted value of R for Solar circle objects by recalculat-ing Ω for a range of values of 7 ≤ R ≤ = ± . − kpc − . The value of Ω is consistent with other Solarcircle and tangent point sources as discussed in detail inBurns et al. (2014). All such objects produce a value thatis higher than that obtained from the ratio of the Galacticconstants recommended by the IAU of Ω = Θ /R = 25 . − kpc − (Kerr & Lynden-Bell, 1986).
5. Conclusions
We measured the trigonometric distance of IRAS20056+3350 to be D = 4 . +0 . − . kpc, which is 2.8 timeslarger than the near D kin often adopted in the literature.In measuring the annual parallax the astrometric accu-racy in the R.A. and Dec. directions were investigatedand the former was found to be significantly influencedby the elongation of maser structure, although the latterwas affected negligibly. The astrometric errors evaluatedin this analysis were well matched by the error floorsrequired to produce a χ value of unity in the parallaxfitting procedure.We determined the systematic proper motion of thesource to be ( µ α cos δ , µ δ ) = ( − . ± . − . ± . − under the assumption that the H O masersused in our observations are associated with a line ofsight bipolar outflow. This assumption is justified bythe spatial and kinematic proximity of masers to theblue lobe of the bipolar outflow as revealed from archive0 R. A. Burns et al. [Vol. ,maps, velocity spectra and an SED model compiled usinginfrared photometry data interpreted using the fittingtool of Robitaille et al. (2007).From the accumulation, re-evaluation and summary ofvarious archive data we find IRAS 20056+3350 to bea young MSFR which is forming a central object ofconsiderable mass ( > M ⊙ ). It is at an evolutionarystage resembling a Class I
YSO and appears to precedethe formation of an UCHII region.IRAS 20056+3350 is at the leading tip of the Local arm,near the Solar circle, and its position in relation to thespiral arms tentatively supports that the Local arm maybe a branch of the Perseus arm.Using our results obtained with VERA, and the specialgeometry applicable to Solar circle objects, we evalu-ated the angular velocity of Galactic rotation at the Sun,Ω = 29 . ± .