Propagation of Highly Efficient Star Formation in NGC 7000
Hideyuki Toujima, Takumi Nagayama, Toshihiro Omodaka, Toshihiro Handa, Yasuhiro Koyama, Hideyuki Kobayashi
aa r X i v : . [ a s t r o - ph . GA ] J u l PASJ:
Publ. Astron. Soc. Japan , 1– ?? , c (cid:13) Propagation of Highly Efficient Star Formation in NGC 7000
Hideyuki
Toujima , Takumi
Nagayama , Toshihiro
Omodaka ,
1, 3
Toshihiro
Handa , Yasuhiro
Koyama , and Hideyuki Kobayashi Graduate School of Science and Engineering, Kagoshima University,1-21-35 Korimoto, Kagoshima, Kagoshima 890-0065 Mizusawa VLBI Observatory, National Astronomical Observatory of Japan,2-21-1 Osawa, Mitaka, Tokyo 181-8588 Faculty of Science, Kagoshima University,1-21-35 Korimoto, Kagoshima, Kagoshima 890-0065 Institute of Astronomy, The Universe of Tokyo,2-21-1 Osawa, Mitaka, Tokyo 181-0015 Kashima Space Research Center, National Institute of Information and Communications Technology,893-1 Hirai, Kashima, Ibaraki [email protected] (Received 2009 October 6; accepted 2011 July 11)
Abstract
We surveyed the (1,1), (2,2), and (3,3) lines of NH and the H O maser toward the molecular cloudL935 in the extended H II region NGC 7000 with an angular resolution of 1 . ′ emission with a size of 0.2–1 pc and mass of 9–452 M ⊙ .The molecular gas in these clumps has a similar gas kinetic temperature of 11–15 K and a line width of1–2 km s − . However, they have different star formation activities such as the concentration of T-Tauritype stars and the association of H O maser sources. We found that these star formation activities arerelated to the geometry of the H II region. The clump associated with the T-Tauri type star cluster has ahigh star formation efficiency of 36–62%. This clump is located near the boundary of the H II region andmolecular cloud. Therefore, we suggest that the star formation efficiency increases because of the triggeredstar formation. Key words:
Star: formation - ISM: H II region - ISM: individual (SFR) - line (NH )
1. Introduction
The star formation efficiencies (SFEs) of molecularclouds in the Milky Way Galaxy, are typically observedto be 10%. The SFEs in nearby molecular clouds are ≃ ≃ ≃ ∗ NGC 7000 is an extended H II region in the CygnusX region. On its southeastern side is a molecular cloudL935. The CO emission in this molecular cloud is thebrightest in the Cygnus X region (Dobashi et al. 1994). ∗ Present address: Graduate School of Science and Engineering,Kagoshima University, 1-21-35 Korimoto, Kagoshima,Kagoshima 890-0065.
Figure 1 shows the optical image of NGC 7000 and L935.Seven T-Tauri stars are clearly clustered at the boundaryof the H II region (Herbig 1958). This suggests that thestar formation is triggered by the interaction of the H II region with the dense molecular gas. A number of studiesof star formation in a cloud associated with an H II re-gion have been performed (Sugitani et al. 1989; Sugitaniet al. 1991; Sugitani & Ogura 1994; Dobashi et al. 2001Deharveng et al. 2003; Deharveng et al. 2005). For exam-ple, in the nearby H II region, IC 5070, a molecular shellwith an expanding velocity of 5 km s − , is found in the CO ( J =1-0) line (Bally & Scoville 1980). They suggestthat the T-Tauri type stars in IC 5070 are formed by theexpanding shell.Our aim is to investigate the relationship between densemolecular gas and star formation based on the SFE.NGC700 has the advantage of allowing us to estimate theSFE because T-Tauri type stars are associated with it andwe can estimate the stellar mass accurately. Therefore,we made the observations in the NH line to estimate themass of dense molecular gas. We also surveyed an H Omaser source that is associated with outflow from a youngstellar object (YSO). We adopted the distance to NGC7000 to be 600 pc (Laugalys & Straiˇzys 2002). H. Toujima et al. [Vol. ,
2. Observations Observations
We observed NGC 7000 in the NH lines with theKashima 34-m telescope of the National institute ofInformation and Communications Technology (NiCT)from April 2007 to October 2008. We made simulta-neous observations in three inversion transitions of theNH ( J, K ) = (1,1), (2,2), and (3,3) lines at 23.694495,23.722633, and 23.870129 GHz, respectively. At 23 GHz,the telescope beam size was 1 . ′ η MB ) was 0.50. We used a K -band HEMTamplifier whose system noise temperature was 150–250K. The relative pointing error was better than 0 . ′
2; thiswas checked by the observations of several H O masersources at 22.235080 GHz. All spectra were obtainedwith an 8192-channel FX-type spectrometer developed atKagoshima University and NiCT. Its bandwidth and fre-quency resolution are 256 MHz and 31.25 kHz, respec-tively. The corresponding velocity coverage and velocityresolutions are 3200 km s − and 0.39 km s − at the NH frequencies, respectively. The total number of the ob-served positions is 311 and the surveyed area is approxi-mately 38 ′ × ′ or 6.6 pc × profiles wereobtained at 1 ′ grid points of the equatorial coordinates.All data were obtained with the position switch betweenthe target and a reference position. The reference positionis ( α,δ ) (J2000) = (20 h m . s , +44 ◦ ′ ′′ ), where no NH emission was detected. We integrated at least 20 min ateach point. The rms noise level was typically 0.20 K inthe unit of the main beam brightness temperature definedby T MB ≡ T ∗ A /η MB , where T ∗ A is the antenna temperaturecalibrated by the chopper wheel method (Kutner & Ulich1981).Data reduction was performed using the UltraSTARpackage developed by the radio astronomy group at theUniversity of Tokyo (Nakajima et al. 2007). In this paper,the intensities are presented in the main beam tempera-ture. O Maser Observations
The single-dish observations of the 6 → transi-tion of the H O maser at 22.235080 GHz were made atthe positions of the NH emission peaks. The observa-tions were made with the Kashima 34-m telescope and theVLBI Exploration of Radio Astrometry (VERA) Iriki 20-m telescope. In the observations made with the Kashima34-m telescope, the conversion factor from the antennatemperature to the flux density was 8 Jy K − for a pointsource. The rms noise level of the H O maser spectra wasless than 1 Jy after integration of 10–30 min. The veloc-ity resolution of the spectrometer was 0.42 km s − at theH O maser frequency. In the observations made with theVERA Iriki 20-m telescope, the conversion factor from theantenna temperature to the flux density was 20 Jy K − .The rms noise level of the H O maser spectra was lessthan 1 Jy after integration of 30 min. The velocity res-olution of the spectrometer was 0.21 km s − at the H Omaser frequency. The VLBI observations of the 6 → transition ofthe H O maser were made with the VERA of the NationalAstronomical Observatory of Japan (NAOJ) on February15, 2008. The data were recorded using the VSOP termi-nal at a rate of 128 Mbit s − . The recorded signals werecorrelated using the Mitaka FX correlator. The spectralresolution was set at 31.25 kHz, which corresponds to avelocity resolution of 0.42 km s − . The system noise tem-perature is approximately 140 K. The phase calibrator wasBL Lac (ICRF J220243.2+421639). The phase-trackingcenter of the array was set at α (J2000) = 20 h m . s δ (J2000) = +43 ◦ ′ . ′′
60. We calibrated the data inthe standard reduction procedure with the AstronomicalImage Processing System (AIPS) of the National RadioAstronomy Observatory (NRAO). The resultant rms noiselevel and synthesized beam size are ≃ − and ≃ × − ◦ .
3. Results Clumps
We observed 311 positions in the survey. NH (1,1)and (2,2) lines were detected with a signal-to-noise ratiogreater than 3 at 138 and 32 positions, respectively. The(3,3) line was not detected at any positions in the ob-served area. We show the profiles towards emission peaksin Figure 2.Figure 3 shows the velocity integrated map of the mainhyperfine component of the NH (1,1) line over v LSR = − − . The NH (1,1) emission extends over 38 ′ × ′ or 6.6 pc × clump as an isolated feature in the (1,1)integrated intensity distribution with intensities strongerthan the 3 σ noise level ( > ∼ . − ) both in the (1,1)and the (2,2) lines. Based on this definition, we identifytwo NH clumps; we call them clump-A at the northeastand clump-B at the southwest. We found local peaks inthese clumps. In clump-A, there are YSOs that may havebeen formed near three local peaks (see subsection 4.1).In order to investigate the star formation of clump-A indetail, we define three subclumps as three local peaks inclump-A. We call them subclump-A1, subclump-A2, andsubclump-A3 from the east to the west. The parametersof the clumps and subclumps are summarized in Table1. As shown in Figure 3, the shape of these subclumpsappears to be an ellipse with the major axis along theright ascension and the minor axis along the declination.Therefore, we estimate the subclump sizes of the majorand minor axes using the FWHM of the Gaussian fittingto the intensities along the right ascension and declination,respectively. Figure 4 shows the results of the Gaussianfitting. This size is used to estimate the mass of eachsubclump.Clump-A extends over 11 ′ × ′ or 1.9 pc × v LSR = 0 to8 km s − are shown in Figure 5. Clump-A appears at v LSR = 5–7 km s − and clump-B appears at v LSR = 1–3km s − .o. ] NH and H O maser emissions in NGC 7000 3Figure 6 shows the position-velocity diagram along thedashed line shown in Figure 3. Clump-A and clump-B areclearly separated in the position-velocity space. It sug-gests that these clumps are not parts of a single object.However, clump-A, clump-B, and other weak features be-tween the clumps are aligned on the position-velocity di-agram. This suggests that these clumps may be formedfrom a single system (see subsection 4.3).A velocity gradient is found in subclump-A3. Figure 7shows the position-velocity diagram of subclump-A3. Theestimated velocity gradient is 1.66 ± − pc − .The linewidth of the peak spectrum is 2.11 ± − , and it is 1.4 times broader than the other spectra inclump-A. The other subclumps do not show the velocitygradient.Clump-B extends over 15 ′ × ′ or 2.6 pc × v LSR = 1–4 km s − (Figure 5). A velocity gradient isfound in the emission peak of clump-B. Figure 7 showsthe position-velocity diagram of the peak of clump-B. Theestimated velocity gradient is − . ± .
23 km s − pc − .The CO ( J =1-0) emission was detected by Dobashiet al. (1994) with a velocity range of v LSR = 1.5–7.5 kms − at the mid position of clump-A and clump-B. It com-prises two velocity components of clump-A and clump-B.The spatial distribution in the NH line is similar to thatin CO line, although the CO observations were madewith a lower angular resolution (2 . ′
7) and sparsed sam-pling (5 ′ ). Clump-B is located at the CO peak both onthe sky and in the velocity. Clump-A is located on thenortheast side of the CO cloud; it faces the H II region. We estimated the physical parameters of the clumpsand the subclumps. NH is a very well-studied moleculewith which to investigate the physical conditions of thedense molecular gas.The NH lines are split by the quadrupole hyperfineinteraction. Optical depths can be directly determinedfrom the intensity ratio of the main to the satellite lines.Because we detect the hyperfine structure in the (1,1) line,the optical depth, τ (1 , , can be derived from the intensityratio. Figure 8 shows the correlation of the integrated in-tensities of the main and the satellite lines. We estimatedtwo intensity ratios of the inner and outer satellite linesto the main line. The optical depths estimated from thesetwo ratios are the same within the error. We list the op-tical depths of the clumps and the subclumps in Table 2,and these are found to be in the range of 0.8–1.6.We estimated the NH rotational temperature from theintensity ratio of the (2,2) line to the (1,1) line using themethod shown by Ho & Townes (1983). Figure 9 showsthe correlations of the integrated intensities in the (1,1)and (2,2) lines. The rotational temperatures of all clumpsand subclumps are 11–15 K and the same within the er-ror. Using the collisional excitation model (Walmsley &Ungerechts 1983; Danby et al. 1988), the rotational tem-perature is estimated to be very close to the gas kinetictemperature, T kin , for T rot <
15 K. Therefore, T kin should be approximately 13 K. The estimated temperatures arelisted in Table 2.We derive the total column density of NH , N (NH ),from the column density in the (1,1) line, assuming thelocal thermodynamic equilibrium (LTE) condition for themolecules in the clumps (Rohlfs & Wilson 1996). Theestimated total column densities of the clumps are listedin Table 2 and they are in the range of N (NH )=(1.8–3.7) × cm − .We derived the molecular gas mass for each of theclumps and subclumps using two methods. One is theLTE mass that is derived from the deconvolved size andthe column density with the assumed abundance ratio.Using a model of a uniform density share with 40% he-lium in mass, the LTE mass is given by M LTE = 467 (cid:18) R [pc] (cid:19) (cid:18) N (NH )10 [cm − ] (cid:19) (cid:18) X (NH )10 − (cid:19) − M ⊙ , (1)where R is the radius of the sphere, and X (NH ) is theabundance of NH relative to H . For clump-A and B, R is given from a geometrical mean of the major and minoraxes of an apparent ellipse after beam deconvolution. Forsubclumps A1–A3, R is given from a geometrical mean ofFWHMs of the Gaussian fitting along the right ascensionand declination. The estimated sizes in diameter are inthe range of 0.6–1 pc for the clumps and 0.2–0.3 pc forthe subclumps.Ho & Townes (1983) reviewed that the abundance ofNH relative to H has been estimated to range from 10 − in the core of the dark cloud L183 (Ungerechts et al. 1980)up to 10 − in the hot core of the Orion KL (Genzel et al.1982) and the ion-molecule chemistry produces an abun-dance of the order of 10 − (Prasad & Huntress 1980). Weuse the abundance ratio of 10 − , because the estimatedkinetic temperature and the clump size in NGC 7000 areclose to those of L183. Using this abundance, the hy-drogen column density derived from NH is derived tobe N (H )=(1.8–3.7) × cm − . This corresponds to A V ≃ N (H ) /A V = 1 . × atoms cm − mag − (Bohlin et al. 1978). Comer´on & Pasquali (2005) esti-mated the extinction of the molecular cloud to be A V ≃ A V are consistent within a factor. Moreover, the hydro-gen column density estimated from CO (Dobashi et al.1994) is consistent with our estimation within a factor of2–3, although the observation grid is different. The LTEmasses of clump-A and clump-B are estimated to be 95and 452 M ⊙ , respectively. The LTE masses of subclump-A1, subclump-A2, and subclump-A3 are estimated to be12, 20, and 9 M ⊙ , respectively.The other mass estimation is the virial mass, M vir . Thisis calculated as M vir = kR ∆ v M ⊙ , where k is taken as 210based on a uniform density sphere (MacLaren et al. 1988), R is the radius of the clump, and ∆ v is the half-power linewidth. The virial masses of clump-A and clump-B areestimated to be 125 and 350 M ⊙ , respectively. The virialmasses of subclump-A1, subclump-A2, and subclump-A3are not estimated, because their sizes are small. H. Toujima et al. [Vol. ,We found that the LTE mass and the virial mass areconsistent by a factor of 1.3. From the above mentionedcomparisons such as A V , CO, and derived mass, wecould estimate the actual masses of molecular gas withina factor of 2–3. All these derived parameters are summa-rized in Table 2. O Maser Source in Subclump-A2
We conducted an H O maser survey of the NH clumpsand subclumps. We discovered a new H O maser emissionat the NH peak in subclump-A2. However, no H O maseremission was detected in the others with the upper limitof 3 Jy (Figure 3).We conducted monitoring observations of the masersource from August 2007 to May 2008. The obtainedspectra are shown in Figure 10. All spectra show a sin-gle velocity component with a narrow linewidth (FWHM)of ≃ − . This maser emission is time-variable be-tween 107 Jy in December 2007 and 5 Jy in May 2008.We found that the maser emission disappeared with theupper limit of 2.1 Jy in October 2008. The LSR velocityof the maser jumped from 8.4 to 9.0 km s − during theperiod from November 2007 to December 2007, and from9.0 to 9.6 km s − from January 2008 to April 2008. Thismeans that there were two maser spots and the lifetimeof each spot is less than a year.Claussen et al. (1996) reported that an H O maser asso-ciated with a low-mass star is variable on a timescale rang-ing from months to a year. The H O maser in subclump-A2 shows variations on the same timescale. Therefore, theH O maser should be associated with low-mass stars.The detected H O maser of v LSR ≃ − is red-shifted with respect to the ambient gas velocity of ≃ − observed in the NH line. This suggests that themaser may be associated with outflows from the protostar.To identify the counterpart of the maser, we should ob-tain an accurate position of the maser. Therefore, weconducted VLBI observations on February 15, 2008, todetermine its position. A single feature with a size of1.6 mas × × α,δ ) J2000 = (20 h m . s , +43 ◦ ′ . ′′
5) by a fringe rateanalysis. The position uncertainty was approximately0 . ′′
1. However, we can find no visible star or optical fea-ture suggesting an outflow at the position of the maserin the DSS2 images. We discuss the counterpart of themaser in subsection 4.1.
4. Discussion
To investigate the star formation in our identified NH clumps, we examined the distribution of young stars re-ported in the previous observations. In clump-A, Herbig(1958) found seven emission-line stars, and Cohen & Kuhi(1979) confirmed that these stars are T-Tauri type stars.T-Tauri type stars are suitable for the mass estimationbecause their mass can be well estimated using the H-R diagram (Cohen & Kuhi 1979). We estimate the mean andtotal masses of T-Tauri type stars in clump-A to be 1.3and 9.0 M ⊙ , respectively. We consider that these T-Tauritype stars indicate the lower limit of the star formation ac-tivity. We found 32 and 11 infrared sources listed in the2MASS and MSX catalogues, respectively, in clump-A.We consider that these infrared sources indicate the upperlimit of the star formation activity, although some of thesesources may be the fore/background sources. There is alarge difference in the distribution of these T-Tauri typestars and infrared sources in each subclump, suggestingdifferences in star formation activity.We found that five of seven T-Tauri type stars are con-centrated in subclump-A1. The total mass of these T-Tauri stars is 6.9 M ⊙ . The concentration of the T-Tauritype stars suggests that the star formation of subclump-A1 is the most active. There are 15 2MASS sources and3 MSX sources in subclump-A1. The T-Tauri type starsand the majority of 2MASS sources are found on the eastside of subclump-A1, where the H II region is located.We found a new H O maser source in subclump-A2.Furuya et al. (2003) reported that ≃
40% of class 0, ≃
4% of class I, and no class II low-mass protostars emit theH O maser. We found the counterpart of the H O maserin the
Spitzer infrared images at 24 and 70 µ m (Figure11). The 70 µ m emission of the counterpart is centeredat ( α, δ ) J2000 = (20 h m . s , +43 ◦ ′ . ′′
9) and extendsover a 20 × ′′ (12000 × O maserobtained by our VLBI observations. The
Spitzer sourceshould be a protostar associated with the H O maser. Weshow a spectrum of the
Spitzer source at six wavelengthsin Figure 12. At the wavelengths of 1.25–8.28 µ m, thecounterpart is not detected with 2MASS and MSX, andwe show the upper limits. A single-temperature black-body radiation through the data points at 24 and 70 µ mis consistent with the upper limits. Its bolometric lumi-nosity is estimated to be 42 L ⊙ . Both the spectrum shapeand the bolometric luminosity are consistent with those ofa class 0 protostar (Bachiller 1996). Therefore, the coun-terpart of the H O maser should be a class 0 protostar. Toconfirm this, the submillimeter observations are required.One T-Tauri type star and 5 2MASS sources arefound in subclump-A2. A nebulosity at ( α, δ ) J2000 =(20 h m s , +43 ◦ ′ ′′ ) is visible in the DSS2 images inthe B , R , and I -bands. This is considered to be a re-flection nebula, because it is continuously detected at theoptical wavelength.T-Tauri type stars are not found in subclump-A3. Four2MASS and four MSX sources are found. These MSXsources would be the YSOs embedded in the dust enve-lope because they are invisible in the DSS2 and 2MASSimages. G085.0482-01.1330 identified by MSX is locatedat the peak position of the (1,1) line in subclump-A3.We found a velocity gradient at this position (see subsec-tion 3.1). The velocity gradient would be due to the sim-ple core rotation or the outflow from G085.0482-01.1330.This source is located at the center of the velocity gra-dient. The spectrum of this source, shown in Figure 12,o. ] NH and H O maser emissions in NGC 7000 5is similar to that of a class I protostar. The bolometricluminosity was estimated to be 190 L ⊙ .T-Tauri type stars and H O maser sources are not foundin clump-B. We found 24 2MASS and 6 MSX sources. Oneof the MSX sources identified as G084.8235-01.1094, islocated at the peak position in the (1,1) line. The velocitygradient is found in the (1,1) line at this position. Thespectrum of this source, shown in Figure 12, is similar tothat of a class I protostar. Its bolometric luminosity wasestimated to be 230 L ⊙ .We consider that the star formation of clump-A is moreactive than that of clump-B. In clump-A, both the numberand the total mass of stars in subclump-A1 are larger thanthose of subclump-A2 and A3. We have presented the stellar mass of each clump inthe previous subsection. In this subsection, we exam-ine the relation between the stellar and the molecular gasmasses. We estimate the star formation efficiency givenby SFE= M star / ( M star + M gas ).The SFE of subclump-A1 is estimated to be ≃
36% fromthe stellar mass of 6.9 M ⊙ and gas mass of 12 M ⊙ . Weuse the stellar mass of the identified T-Tauri type stars inthis estimation. However, other T-Tauri type stars may beassociated with the NH clumps but located just behindthem. In this case, these T-Tauri type stars could be de-tected in the K -band of 2MASS. We obtained the extinc-tion in the K -band to be A K ≃ A V ≃
10 magestimated from the NH column density and the extinc-tion law of A K / A V = 0.112 (Rieke & Lebofsky 1985). Forthe identified T-Tauri type stars, the K -band magnitudeof 2MASS is 8.5–11.5 mag. Therefore, the K -band mag-nitude of a T-Tauri type star located behind the clumps isestimated to be 9.5–12.5 mag. This value is brighter thanthe 2MASS K -band detection limit of 14.3 mag (SNR =10). We found ten 2MASS sources that are not identifiedas T-Tauri type stars in subclump-A1. In the case thatall of them are T-Tauri type stars with a mass of 1.3 M ⊙ ,the SFE of subclump-A1 increases to ≃ M ⊙ is found. Therefore, its SFE is estimated to be ≃
4% from the gas mass of 20 M ⊙ . The SFE averagedover the whole clump-A is estimated to be ≃
8% fromthe stellar mass of 7.7 M ⊙ and gas mass of 95 M ⊙ . TheSFEs of both subclump-A3 and clump-B might be 0%because no T-Tauri type star is found there. In the casethat the 2MASS sources are included in the stellar massestimation, the SFEs of subclump-A2, A3, and the wholeof clump-A are estimated to be 23–36%. This value is closeto the SFE of subclump-A1. However, clump-B showslower SFE of 6% even in this case.In either case, including only T-Tauri type stars or alsothe 2 MASS sources, the SFE of subclump-A1 is estimatedto be 36–62%; this is higher than the SFEs of the othermolecular clouds. In order to make a fair comparison,we revisited the SFEs of the following three molecularclouds using the same procedure. The SFE of NGC 2264is estimated to be 11% from the stellar mass of the OB and T-Tauri type stars of 119 M ⊙ (Dahm & Simon 2005)and the molecular gas mass traced in the NH line of1000 M ⊙ (Lang & Willson 1980). The SFE of NGC 1333is estimated to be 16% from the stellar mass of the T-Tauri type stars of 17 M ⊙ (Aspin 2003) and the moleculargas mass traced in the NH line of 106 M ⊙ (Ladd et al.1994). The SFE of L1228 is estimated to be 8% from thestellar mass of the T-Tauri type stars of 1 M ⊙ (Kun et al.2009) and the molecular gas mass traced in the NH lineof 12 M ⊙ (Anglada et al. 1994). The SFE of subclump-A1 is close to ≃
42% estimated at NGC 2024 and NGC2068 from the CS observations (Lada 1992). The derivedSFEs of individual clumps are summarized in Table 3.The values of “Total” in Table 3 correspond to the upperlimits of the SFEs for the case in which the all 2MASSsources we found are associated with the clump or thesubclump.
Although clump-A and clump-B are adjacent on thesky, there is a big difference in the star formation activ-ities. Because these clumps are close to NGC 7000, thedifference may be due to the H II region. Therefore, wediscuss the geometry of the clumps and the H II region.The optical image (Figure 1) shows that the clumps arelocated in the foreground of the H II region. Subclump-A1is the nearest to and clump-B is the farthest from the H II region on the sky.To investigate the three-dimensional structure of clump-A, we estimated the length along the line of sight, l , de-rived as l = N (H ) /n cr , where N (H ) and n cr are the hy-drogen column density and the critical density in the NH line, respectively. When we use n cr ≃ cm − (Myers& Benson 1983), the lengths of subclump-A1, A2, and A3are estimated to be l ≃ M ⊙ . If the two clumps are gravitation-ally bound, the enclosed mass is estimated to be 11000 M ⊙ from their separation of 2.9 pc and relative velocity of4 km s − . The LTE mass is ∼ CO line is estimated to be 3400 M ⊙ from the columndensity of 1 . × cm − (Dobashi et al. 1994). Boththe total mass of the two clumps and the CO cloud issmaller than the enclosed mass. Therefore, clump-A andclump-B are gravitationally unbound. However, clump-A, clump-B, and other weak features between the clumpsare aligned on the position-velocity diagram (Figure 6).This suggests that they are also aligned in the three-dimensional structure.We compare the LSR velocities of the H II region and themolecular gas. Figure 13(a) shows the LSR velocity mapof the H α emission (Fountain et al. 1983) superimposed onthe CO integrated intensity map (Dobashi et al. 1994) H. Toujima et al. [Vol. ,and the NH (1,1) integrated intensity map. The LSR ve-locity of the H α emission line is approximately 0 km s − atthe position overlapped with the NH clumps, and 4–5 kms − on both the eastern and the western sides of the NH clumps. The LSR velocities of clump-A and clump-B are5.5 and 1.5 km s − , respectively. Clump-A is redshiftedwith respect to the H II region. We show the FWHM mapof the H α emission (Fountain et al. 1983) in Figure 13(b).The FWHM of the H α emission around the NH clumpsis 10–20 km s − . This value is narrower than the FWHMat the other position. These characteristics can be inter-preted as indicating that the redshifted component of theH α emission is blocked by clump-A, and only ionized gaslocated in the foreground of clump-A would be observed.Therefore, clump-A would be surrounded by the ionizedgas.In clump-B, the H α emission is not detected. This in-dicates that there is no ionized gas at the foreground ofclump-B. However, we found the presence of ionized gasin the background of clump-B, because a radio continuumemission at 4.8 GHz is detected there (Wendker 1984).Therefore, clump-B would be located in the foreground ofthe H II region. We show the schematic geometry of theNH clumps and the H II region in Figure 14. In subsection 4.2, we show that the SFE ≃ II region. The ge-ometry shown in the previous subsection suggests thatsubclump-A1 is closer to the H II region than any of theother subclumps or clump-B. The five T-Tauri stars insubclump-A1 would be formed by the interaction of theH II region with the molecular gas. A high SFE is ob-served in other triggered star forming regions. The NGC2024 and 2068 molecular clouds, each of which interactwith an H II region (Chandler & Carlstrom 1996), showSFE ≃
42% (Lada 1992). This means that the SFE ofsubclump-A1 increases because of the effect of the H II region.Theoretical calculations suggest that the SFE is in therange of 30–50% of the regions that form clusters of low-mass stars (Matzner & McKee 2000). The estimated SFEof subclump-A1 is very close to this value. Matzner &McKee (2000) suggest that the SFE of the clumps thatare more massive than approximately 3000 M ⊙ , in whichO stars will form, is lower than 30–50% because of thedestructive effects of massive stars. The SFE may be in-creased in a cloud with the formation of a low-mass starcluster.We consider whether other subclumps and clump-B arekept the low SFE. Star formation appears to advance se-quentially in the order of A1, A2, A3, and B. This orderis the same as that of the distance from the H II region. It is suggested that the triggered star formation or the inter-stellar shock comes sequentially from the H II region. Amolecular shell with an expansion velocity of ≃ − isfound in the CO line (Bally & Scoville 1980). In the casethat the effect of the H II region expands at this velocity,the crossing timescale from subclump-A1 to clump-B isestimated to be ≃ × yr using their separation of 2.9pc. This timescale is shorter than the lifetime of the O5V type star (2 × yr; Walborn 2007), which is consid-ered to be the ionizing star of the H II region (Comer´on &Pasquali 2005). This suggests that the H II region can af-fect clump-B in the future, in the case that the separationin the line of sight is the same order of magnitude as thaton the sky. The total molecular gas mass of clump-A andclump-B is 391 M ⊙ . This mass is close to that of NGC2024 or NGC 2068 (Lada 1992), and it is small enoughto avoid cloud destruction by new born stars (Matzner &McKee 2000). This suggests that the SFE of the combinedclump-A and B can be as high as approximately 40%.The observed SFE is sensitive to the estimation of bothgas and stellar masses. As mentioned in subsection 3.2,we estimated the actual gas mass within a factor of 2–3.However, it is generally difficult to estimate the moleculargas mass. The estimated molecular gas mass is sometimesdifferent by more than a factor of 10 in different observedlines. In the W3 giant molecular cloud, the moleculargas mass estimated in the CO ( J =1-0), C O ( J =2-1),and NH lines is 16000, 1400, and 3300 M ⊙ , respectively(Tieftrunk et al. 1998), although they were estimated inthe same area. For the estimation of SFE, the CO linedata are often used to estimate the molecular gas mass(e.g. Myers et al. 1986; Leisawitz et al. 1989). Becausethe CO line traces the less dense gas, the molecular gasmass might be overestimated. However, the data of themolecular line to trace the dense gas is not ideal to esti-mate the molecular gas mass, because the relative abun-dance is difficult to determine precisely. For example, theabundance of NH varies by up to a factor of 10 fromcloud to cloud.It is also difficult to estimate the stellar mass accurately.There are few studies based on total stellar mass estimatedas the sum of the masses of individual stars. Althoughthe infrared luminosity is often used to estimate the stel-lar mass, it would be less accurate than the mass esti-mated based on the number count of the T-Tauri typestars shown in this paper.As mentioned above, the observed SFE reported insome studies should be revised by an order of magnitude.The SFEs of the nearby molecular clouds such as Perseusand Ophiuchus, L1551 in Taurus, and giant molecularclouds in the inner Galaxy are 3–6%, 9–15%, and 2%, re-spectively (Myers et al. 1986; Swift & Welch 2008; Evanset al. 2009). These values may be underestimated, be-cause these studies use the molecular gas mass estimatedfrom the A V and CO maps. We should take care how toestimate the SFE to refer it from the previous studies.o. ] NH and H O maser emissions in NGC 7000 7 O Maser Surveys
Previous surveys of H O masers have been carried outbased on the IRAS Point Source Catalogue (PSC). Thiscatalogue is useful for searching for YSOs embedded in themolecular clouds. However, the counterpart of the H Omaser that we found in subclump-A2 is not catalogued.It shows that the H O survey based on the IRAS PSC isinsufficient. There are two possibilities why some YSOsare uncatalogued in the IRAS PSC: sensitivity too poorto detect them or resolution too poor to resolve a clusterof some sources. Figure 15 shows an IRAS image at 100 µ m overlayed on our NH map. A complex source is foundnear clump-A in the IRAS image, although it is composedof several infrared sources in the Spitzer image (see Figure11).This suggests that there are many H O maser sourceswhich are not catalogued in the IRAS PSC. A new H Omaser survey should be carried out based on a point sourcecatalogue with a higher resolution and sensitivity, such asq
Spitzer and/or AKARI should be carried out. Our newH O maser is associated with a far-infrared source, andits luminosity is brighter than that of the mid-infrared.This characteristic may be a good criterion with which tofind new H O maser sources.
5. Conclusions
We observed NGC 7000 in the NH line and H O maserusing the Kashima 34-m telescope. Our observations aresummarized as follows:1. We found two major clumps with a mass of 95–452 M ⊙ , and three subclumps with a mass of 9–20 M ⊙ .The molecular gas in these show similar gas kinetictemperatures of 11–15 K and line width of 1–2 kms − . However, they show different star formationactivities such as the concentration of T-Tauri typestars and the association of an H O maser.2. One of the clumps that is associated with a cluster ofT-Tauri type stars shows the SFE ≃ α emission suggeststhat the clump with high SFE is located near the H II region. Therefore, the high SFE would be related tothe interaction of molecular gas and the H II region.4. We found a new H O maser source in the NH clump. Although the counterpart of this maser isnot found in the IRAS point source catalogue, wefound it in the Spitzer
24- and 70- µ m images. Thissuggests that a new H O maser survey should becarried out based on the point source catalogue of
Spitzer and/or AKARI.We thank an anonymous referee for very useful com-ments and suggestions. T.O. was supported by a Grant-in-Aid for Scientific Research from the Japan Society forthe Promotion Science (17340055). We acknowledge K. Miyazawa (NAOJ) for his technical support of observa-tions.
References
Anglada, G., Rodriguez, L. F., Girart, J. M., Estalella, R., &Torrelles, J. M. 1994, ApJL, 420, L91Aspin, C. 2003, AJ, 125, 1480Bachiller, R. 1996, ARA&A, 34, 111Bally, J., & Scoville, N. Z. 1980, ApJ, 239, 121Bohlin, R. C., Savage, B. D., & Drake, J. F. 1978, ApJ, 224,132Cambr´esy, L., Beichman, C. A., Jarrett, T. H., & Cutri, R. M.2002, AJ, 123, 2559Chandler, C. J., & Carlstrom, J. E. 1996, ApJ, 466, 338Claussen, M. J., Wilking, B. A., Benson, P. J., Wootten, A.,Myers, P. C., & Terebey, S. 1996, ApJS, 106, 111Codella, C., Welser, R., Henkel, C., Benson, P. J., & Myers,P. C. 1997, A&A, 324, 203Cohen, M., & Kuhi, L. V. 1979, ApJS, 41, 743Comer´on, F., & Pasquali, A. 2005, A&A, 430, 541Dahm, S. E., & Simon, T. 2005, AJ, 129, 829Danby, G., Flower, D. R., Valiron, P., Schilke, P., & Walmsley,C. M. 1988, MNRAS, 235, 229Deharveng, L., Lefloch, B., Zavagno, A., Caplan, J.,Whitworth, A. P., Nadeau, D., & Mart´ın, S. 2003, A&A,408, L25Deharveng, L., Zavagno, A., & Caplan, J. 2005, A&A, 433,565Dobashi, K., Bernard, J.-P., Yonekura, Y., & Fukui, Y. 1994,ApJS, 95, 419Dobashi, K., Yonekura, Y., Matsumoto, T., Momose, M., Sato,F., Bernard, J.-P., & Ogawa, H. 2001, PASJ, 53, 85Duerr, R., Imhoff, C. L., & Lada, C. J. 1982, ApJ, 261, 135Evans, N. J., et al. 2009, ApJS, 181, 321Fountain, W. F., Gary, G. A., & Odell, C. R. 1983, ApJ, 269,164Furuya, R. S., Kitamura, Y., Wootten, A., Claussen, M. J., &Kawabe, R. 2003, ApJS, 144, 71Genzel, R., Ho, P. T. P., Bieging, J., & Downes, D. 1982,ApJL, 259, L103Herbig, G. H. 1958, ApJ, 128, 259Ho, P. T. P., & Townes, C. H. 1983, ARA&A, 21, 239Gandolfi, D., et al. 2008, ApJ, 687, 1303Kun, M., Balog, Z., Kenyon, S. J., Mamajek, E. E., &Gutermuth, R. A. 2009, ApJS, 185, 451Kutner, M. L., & Ulich, B. L. 1981, ApJ, 250, 341Lada, E. A. 1992, ApJL, 393, L25Lada, E. A., Evans, N. J., II, & Falgarone, E. 1997, ApJ, 488,286Ladd, E. F., Myers, P. C., & Goodman, A. A. 1994, ApJ, 433,117Laugalys, V., & Straiˇzys, V. 2002, Baltic Astronomy, 11, 205Lang, K. R., & Willson, R. F. 1980, ApJ, 238, 867Leisawitz, D., Bash, F. N., & Thaddeus, P. 1989, ApJS, 70,731MacLaren, I., Richardson, K. M., & Wolfendale, A. W. 1988,ApJ, 333, 821Matzner, C. D., & McKee, C. F. 2000, ApJ, 545, 364Myers, P. C., & Benson, P. J. 1983, ApJ, 266, 309Myers, P. C., Dame, T. M., Thaddeus, P., Cohen, R. S.,Silverberg, R. F., Dwek, E., & Hauser, M. G. 1986, ApJ,301, 398Nakajima, T., et al. 2007, PASJ, 59, 1005
H. Toujima et al. [Vol. ,
Prasad, S. S., & Huntress, W. T., Jr. 1980, ApJS, 43, 1Ridge, N. A., Wilson, T. L., Megeath, S. T., Allen, L. E., &Myers, P. C. 2003, AJ, 126, 286Rieke, G. H., & Lebofsky, M. J. 1985, ApJ, 288, 618Rohlfs, K., & Wilson, T. L. 1996, Tools of Radio Astronomy,XVI, 423 pp. 127 figs., 20 tabs.. Springer-Verlag BerlinHeidelberg New York. Also Astronomy and AstrophysicsLibrary,Straiˇzys, V., & Laugalys, V. 2008, Baltic Astronomy, 17, 143Sugitani, K., Fukui, Y., Mizuni, A., & Ohashi, N. 1989, ApJL,342, L87Sugitani, K., Fukui, Y., & Ogura, K. 1991, ApJS, 77, 59Sugitani, K., & Ogura, K. 1994, ApJS, 92, 163Swift, J. J., & Welch, W. J. 2008, ApJS, 174, 202Tieftrunk, A. R., Megeath, S. T., Wilson, T. L., & Rayner,J. T. 1998, A&A, 336, 991Ungerechts, H., Walmsley, C. M., & Winnewisser, G. 1980,A&A, 88, 259Ungerechts, H., Winnewisser, G., & Walmsley, C. M. 1982,A&A, 111, 339Walborn, N. R. 2007, arXiv:astro-ph/0701573Walmsley, C. M., & Ungerechts, H. 1983, A&A, 122, 164Wendker, H. J. 1984, A&AS, 58, 291 o. ] NH and H O maser emissions in NGC 7000 9
Table 1.
Line parameters obtained at the peak position of the clumps and the subclumps
Clump R.A. Decl. Line T MB ∗ v LSR † ∆ v † R T MB dv ‡ rms noise(J2000) (J2000) ( J, K ) (K) (km s − ) (km s − ) (K km s − ) (K)A1 20 h m . s ◦ ′ ′′ (1,1) 1.92 5.6 1.5 2.96 ± ± ≤ § · · · · · · · · · h m . s ◦ ′ ′′ (1,1) 2.59 5.5 1.4 3.85 ± ± ≤ § · · · · · · · · · h m . s ◦ ′ ′′ (1,1) 1.44 5.3 2.1 3.21 ± ± ≤ § · · · · · · · · · h m . s ◦ ′ ′′ (1,1) 2.79 1.5 1.9 6.25 ± ± ≤ § · · · · · · · · · ∗ The error of the Gaussian fitting is close to the rms noise level. † The error of the Gaussian fitting is much smaller than the velocity resolution (0.39 km s − ). ‡ The error corresponds to one standard deviation. § The upper limit is given as 3 times of the rms noise.
Table 2.
Physical properties of the clumps and the subclumps
Clump Size τ (1 , T rot N (NH ) M LTE M vir (pc) (K) (cm − ) ( M ⊙ ) ( M ⊙ )A1 0.21 1.4 ± ± × · · · A2 0.31 1.2 ± ± × · · · A3 0.21 0.8 ± ± × · · · A 0.67 1.2 ± ± ×
95 125B 1.02 1.6 ± ± ×
452 350
Table 3.
Star formation efficiency of individual clumps
Clump Number of sources Stellar mass ( M ⊙ ) SFE (%)T-Tauri Total T-Tauri Total T-Tauri TotalA1 5 <
15 6.9 < . < < < . < < < . <
36A 7 <
32 7.7 < . <
30B 0 <
24 0 < . <
60 H. Toujima et al. [Vol. ,
Fig. 1.
Optical image of the W80 region (CalTech/Palomar) that comprises two H II regions of NGC 7000 (North Amrica nebula)and IC 5070 (Pelican nebula) and dark lanes of L935. Yellow crosses show the positions of T-Tauri type stars (Herbig 1958). A bluecross shows the position of an O5 V type star (Comer´on & Pasquali 2005). A1(+3,0) A3(-3,0) B(-13,-10) M a i n b e a m t e m p e r a t u r e [ K ] A2(0,0) ( J,K )=(1,1)(
J,K )=(2,2)
LSR velocity [km s -1 ] Fig. 2.
Spectra in the NH ( J, K )=(1,1) and (2,2) lines observed at the peak position of the subclumps and the clump-B. Theposition offsets in arcmin from ( α, δ ) (J2000) = (20 h m . s , +43 ◦ ′ ′′ ) are shown in the top-right corner. o. ] NH and H O maser emissions in NGC 7000 11 l b h m h m o ' o ' RIGHT ASCENSION (J2000) D E C L I NA T I ON ( J ) NGC 7000NH (1,1)beam 1'.6(0.28 pc) AA1 A2 A3 B
Fig. 3.
Integrated intensity map of the main hyperfine component of the NH (1,1) line. The lowest contour and the contourinterval are 0.3 and 0.6 K km s − , respectively. The circle in the bottom-left corner shows the beam size. The gray ellipses showthe extents of the clumps and the subclumps. The triangles and squares show the T-Tauri stars and MSX sources, respectively. Thecross shows the position of the H O maser. The dashed line shows the axis of the position velocity map shown in Figure 6. I n t e g r a t e d i n t e n s it y ( K k m s - ) RA offset (arcmin) A1A2A3A0123 -202 I n t e g r a t e d i n t e n s it y ( K k m s - ) Decl offset (arcmin)A1 01234 -202Decl offset (arcmin)A2 01234 -202Decl offset (arcmin)A3
Fig. 4.
The results of the Gaussian fitting to the intensities along the right ascension and declination for subclump A1–A3. Theposition offsets in arcmin from ( α, δ ) (J2000) = (20 h m . s , +43 ◦ ′ ′′ ) are shown. v LSR = 0 km s − v LSR = 1 km s − v LSR = 4 km s − v LSR = 5 km s − v LSR = 6 km s − v LSR = 7 km s − v LSR = 2 km s − v LSR = 3 km s − NGC 7000NH (1,1)beam 1'.6(0.28 pc)20 h m h m h m h m h m h m h m h m RIGHT ASCENSION (J2000) D E C L I NA T I ON ( J ) o ' o ' o ' o ' Fig. 5.
Velocity channel maps in the NH ( J, K )=(1,1) line. The lowest contour and the contour interval are 0.3 and 0.3 K,respectively. The intensity is averaged in the velocity span of 1 km s − . o. ] NH and H O maser emissions in NGC 7000 13
Position offset [arcmin] L S R v e l o c it y [ k m s − ] +3 0 − − − − − − − A B
Fig. 6.
Position-velocity map of the whole cloud in the NH ( J,K )=(1,1) line. The lowest contour and the contour interval are 0.2and 0.2 K at the T MB unit, respectively. The position offset is relative to the position at ( α, δ ) (J2000) = (20 h m . s , +43 ◦ ′ ′′ )along the dashed line shown in Figure 3. L S R V e l o c i t y [ k m s - ] Position offset [arcmin] -3 Position offset [arcmin] L S R V e l o c i t y [ k m s - ] A3 peak B peak velocity gradient0.32 +/ − − arcmin − − − pc − velocity gradient0.48 +/ − − arcmin − − − pc − Fig. 7.
Position-velocity maps of the NH ( J, K )=(1,1) line toward subclump-A3 (left) and clump-B (right). The lowest contourand the contour interval are 0.3 and 0.3 K, respectively. The position offset is relative to the peak positions at ( α, δ ) (J2000) =(20 h m . s
5, +43 ◦ ′ ′′ ) along the position angle of 45 ◦ for the A3 peak and (20 h m . s
9, +43 ◦ ′ ′′ ) along the position angleof 90 ◦ for the B peak. W m a i n [ K k m s - ] W satellite [K km s -1 ] W m a i n [ K k m s - ] W satellite [K km s -1 ] A1 : inner clump-B : innerclump-A : innerA3 : innerA2 : innerA1 : outer clump-B : outerclump-A : outerA3 :outerA2 :outer
Fig. 8.
Correlations of the integrated intensity in the (1,1) main and satellite lines. The correlations of the main to the inner andouter satellite lines are shown in the top and bottom panel, respectively. The data detected over the 3 σ level in both the main andthe satellite lines are plotted. The error bar shows the rms noise (1 σ ). The estimated optical depth is shown in the bottom of eachpanel. W NH (1,1) [K km s -1 ]
0 1 2 3 4 6 5 0 1 2 3 4 6 5 0 1 2 3 4 6 5 0 1 2 3 4 6 5 0 1 2 3 4 6 5 W NH ( , ) [ K k m s - ] sub-clump A1 R (2,2)/(1,1) = 0.18 +- 0.05 T rot = 12 +- 2 K sub-clump A3 R (2,2)/(1,1) = 0.29 +- 0.05 T rot = 15 +- 2 Ksub-clump A2 R (2,2)/(1,1) = 0.22 +- 0.03 T rot = 13 +- 1 K clump A R (2,2)/(1,1) = 0.24 +- 0.02 T rot = 13 +- 1 K clump B R (2,2)/(1,1) = 0.18 +- 0.02 T rot = 11 +- 1 K Fig. 9.
Correlations of the integrated intensity in the (1,1) and (2,2) lines. The data detected over the 3 σ level in both the (1,1)and the (2,2) lines are plotted. The error bar shows the rms noise (1 σ ). The estimated T rot is shown in each panel. o. ] NH and H O maser emissions in NGC 7000 15 − ]3020100906030060402008060402001208040090603002010010501050 F l ux d e n s it y [ J y ] − − − Fig. 10. H O maser spectra obtained by our single-dish monitoring observations. Two velocity jumps are found between Novemberand December of 2007, and between January and April of 2008. The maser emission disappeared in October 2008.
Fig. 11.
Spitzer images at 24 µ m (left) and 70 µ m (right). A white arrow shows the position of the H O maser. o. ] NH and H O maser emissions in NGC 7000 17
10 -6 10 -5 10 -4 10 -3
Wavelength [m]
10 -7 F l u x d e n s i t y [ J y ]
10 -410 -210 21 class 0class Iclass II counterpart ofthe H O maserin A2 T = 44 K L = L
10 -410 -210 21
G084.8235-01.1094in B T = 130 K L = 230 L
10 -410 -210 21
G085.0482-01.1330in A3 T = 130 K L = 190 L Fig. 12.
Spectral energy distribution of the three infrared sources. The solid line shows the best fit blackbody with parametersshown at the right-bottom corner in each panel. The gray dashed lines in the top panel show typical spectra of class 0, I, and II(Bachiller 1996). o o o o
30' RIGHT ASCENSION (J2000) D E C L I NA T I ON ( J ) LSR velocity [km s − ] − − − (a) o o o o D E C L I NA T I ON ( J )
12 15 27 33 3918 21 24 30 36 42FWHM [km s − ] (b) Fig. 13. (a) The LSR velocity map of the H α emission (color; Fountain et al. 1983) on which the CO integrated intensity map(black contour; Dobashi et al. 1994) and the NH (1,1) integrated intensity map (red contour) are superimposed. The black crossis the emission line star (Herbig 1958). (b) The FWHM map of the H α emission (color; Fountain et al. 1983) superimposed on thesame maps as that shown in (a). o. ] NH and H O maser emissions in NGC 7000 19
Fig. 14.
Schematic geometry of the NH clumps and the H II region. Fig. 15.
IRAS 100 µ m image (gray scale) on which the integrated intensity map of NH3