[email protected] arxiv:1412.0462v1 [astro-ph.ga] 1 ... file1612-mhz line with a double-peak...
TRANSCRIPT
arX
iv:1
412.
0462
v1 [
astr
o-ph
.GA
] 1
Dec
201
4
Astronomy& Astrophysicsmanuscript no. aa23083-13 c©ESO 2018September 2, 2018
OH and H2O maser variations in W33BP. Colom1, E. E. Lekht2, M. I. Pashchenko2, and G. M. Rudnitskij2
1 LESIA, Observatoire de Paris, Section de Meudon, 5 place Jules Janssen, 92195 Meudon CEDEX, Francee-mail:[email protected]
2 Lomonosov Moscow State University, Sternberg Astronomical Institute, 13 Universitetskij prospekt, Moscow, 119234,Russiae-mail:[email protected]
Received....00, 2013; accepted....00, 2014
ABSTRACT
Context. The active star-forming region W33B is a source of OH and H2O maser emission located in distinct zones around the centralobject.Aims. The aim was to obtain the complete Stokes pattern of polarised OH maser emission and to trace its variability and to investigateflares and long-term variability of the H2O maser and evolution of individual emission features.Methods. Observations in the OH lines at a wavelength of 18 cm were carried out on the Nançay radio telescope (France) at anumber of epochs in 2008–2014; H2O line observations (long-term monitoring) atλ = 1.35 cm were performed on the 22-metre radiotelescope of the Pushchino Radio Astronomy Observatory (Russia) between 1981 and 2014.Results. We have observed strong variability of the emission features in the main 1665- and 1667-MHz OH lines as well as in the1612-MHz satellite line. Zeeman splitting has been detected in the 1665-MHz OH line at 62 km s−1 and in the 1667-MHz line at 62and 64 km s−1. The magnetic field intensity was estimated to be from 2 to 3 mG. The H2O emission features form filaments, chainswith radial-velocity gradients, or more complicated structures including large-scale ones.Conclusions. Long-term observations of the hydroxyl maser in the W33B region have revealed narrowband polarised emission in the1612-MHz line with a double-peak profile characteristic of Type IIb circumstellar masers. The 30-year monitoring of thewater-vapourmaser in W33B showed several strong flares of the H2O line. The observed radial-velocity drift of the H2O emission features suggestspropagation of an excitation wave in the masering medium with a gradient of radial velocities. In OH and H2O masers some turbulentmotions of material are inferred.
Key words. masers – ISM:molecules – ISM:radio lines – stars:formation
1. Introduction
Early studies showed that the thermal radio source W33 con-sists of two HII regions: strong and compact G12.80−0.20 anda fainter extended one G12.68−0.18 (see e.g. Goss & Shaver1970). Subsequent observations by Goss et al. (1978) showedthat the extended component (G12.80−0.18) in W33 consistsof several faint discrete features. For the stronger sourceGard-ner et al. (1975) found from H109α, H134α, and H158α ra-dio recombination lines (RRL) radial velocities 35.8, 32.0and38.6 km s−1, respectively. With this velocity the preferable kine-matic distance to the W33 complex was believed to be 4.4 kpc(Haschick & Ho 1983). Quireza et al. (2006) observed towardW33 RRL C91α and C92α. The radial velocity of the extendedregion from the H134α line is 58 km s−1 (Gardner et al. 1975).The accepted model of the W33 region was an interstellar cloudexpanding at a velocity of∼13 km s−1, thus producing in its ra-dio line spectra two Doppler components at approximately 32and 58 km s−1.
Send offprint requests to: P. ColomA complete version of Figure 7 is available in elec-tronic form from http://www.aanda.org/. OH andH2O data in ASCII format are available at the CDSvia anonymous ftp at cdsarc.u-strasbg.fr (130.79.128.5),http://cdsweb.u-strasbg.fr/cgi-bin/qcat?J/A+A/, orhttp://comet.sai.msu.ru/~gmr/Maser_monitoring/W33B/
Pandian et al. (2008) resolved the kinematic distance ambi-guity for a number of galactic HII regions, among them W33,using 21 cm HI absorption spectra. For this purpose they em-ployed a method proposed by Kolpak et al. (2003). From the ab-sence of an HI absorption feature at the tangential-point velocityin the direction of W33 Pandian et al. (2008) concluded that W33is at the near kinematic distance. They estimated its distance as4.9–5.1 kpc with a probable error of±0.6 kpc. Finally, Forster& Caswell (1989) give for W33B a somewhat larger distance of6.4 kpc.
However, the latest measurement of the trigonometric paral-lax of H2O masers in W33B (Immer et al. 2013) yields for thissource a distance of 2.40+0.17
−0.15 kpc, thus placing it in the Scutumspiral arm. It was shown that all the maser sources in the W33region, together with some nearby masers outlining the ScutumArm (Sato et al. 2014), are interconnected and are at the samedistance; the sources A and B possess proper motions with re-spect to C and to their central stars. The proper motions of theH2O masers are discussed below.
Toward W33 three OH and H2O maser emission sources areobserved: W33 A, B, and C. The W33 A and B masers are ar-ranged symmetrically relative to W33C and are at an angulardistance of 7′.5 from it. Sources B and C are associated withthe HII regions G12.68−0.18 and G12.80−0.20, respectively.Source A is at the periphery of the W33 region, near the faintfeature G12.91−0.28 (Goss et al. 1978).
Article number, page 1 of 18
A&A proofs:manuscript no. aa23083-13
Maser emission in the main OH lines 1665 and 1667 MHzwas detected toward W33B by Goss (1968). It was observed in avelocity interval of 58–66 km s−1. Robinson et al. (1970) mademeasurements in the main lines as well as in the 1612- and 1720-MHz satellite OH lines; they managed to detect only the main-line emission. The velocity coincidence of this HII region andthe OH maser source testifies to their physical association.
In 1978 Pashchenko (1980) detected on the Nançay radiotelescope thermal emission and absorption in the satelliteOHlines toward W33B coming from a∼ 7′ × 7′ extended source(molecular cloud) as well as weakly polarised emission fromapointlike source in the 1612-MHz line. Observations in 1978onthe same radio telescope showed that the main-line OH emissionis strongly polarised circularly. The emission (and absorption) ina velocity interval of 30–40 km s−1 belongs to the sources W33Cand A, and in an interval of 55–65 km s−1 to the source W33B.
The H2O maser emission was detected toward W33 byGenzel & Downes (1977) at virtually the same radial veloci-ties as the OH emission. An exception is the emission at smallnegative velocities in W33C. In contrast to OH, a strongerH2O source in this region is W33B. Subsequent observations(Jaffe et al. 1981, and this work) confirmed this characteristicof the H2O masers.
The W33B region also hosts a strong source of maser emis-sion of methanol (CH3OH) in the 51 − 60 A+ 6.67-GHz rota-tional line (Menten 1991). The line profile consists of two peaksat vLSR ∼ 52 and 58 km s−1.
According to the VLA observations of Forster & Caswell(1989, 1999) for most maser sources associated with star-forming regions, OH and H2O masers occur in small groups witha diameter of less than 0.03 pc. They have a common source ofenergy, but arec physically located in distinct zones.
2. Observations and data presentation
We observed the W33B radio source in the 18 cm hydroxyllines at various epochs on the telescope of the Nançay RadioAstronomy Station of the Paris–Meudon Observatory (France).The method of observation and processing of data was presentedby Slysh et al. (2010) and Lekht et al. (2012). At declinationδ = 0◦ the telescope beam at a wavelength of 18 cm is 3.′5× 19′
in right ascension and declination. The telescope sensitivity atλ = 18 cm andδ = 0◦ is 1.4 K/Jy. The system noise tempera-ture of the helium-cooled front-end amplifiers is from 35 to 60 Kdepending on the observational conditions.
We observed H2O maser emission in the 1.35 cm line to-ward W33B (α2000 = 18h13m54.s7, δ2000 = −18◦1′46.′′5) on the22-metre radio telescope in Pushchino from February 1981 toJanuary 2014 with a half-power beamwidth of 2.′6. In the ob-servations of this source the system noise temperature withahelium-cooled field-effect transistor (FET) front-end amplifierwas 120 – 270 K depending on the weather conditions. The sig-nal spectrum was measured by a 128-channel filter-bank anal-yser with a velocity resolution of 0.101 km s−1, and since theend of 2005 by a 2048-channel autocorrelator with a resolutionof 0.0822 km s−1. For a pointlike source an antenna temperatureof 1 K corresponds to a flux density of 25 Jy.
Figure 1 presents the results of observations of hydroxylmaser emission in the 1665- and 1667-MHz lines. The obser-vations in 2008 were carried out with a velocity resolution of0.137 km s−1, and those of 2010–2014 with a resolution of0.068 km s−1. The technique of the observations was describedin detail by Pashchenko et al. (2009) and Slysh et al. (2010).
50 55 60 65 70
0
5
10
15
20March 1, 2014
0
4
8
12
March 8, 2012
0
10
20
LR
Flu
x de
nsity
, Jy
January 6, 2008
1665 MHz
0
4
8
LR
January 6, 20081667 MHz
0
10
20
30April 6, 2010 0
4
8 April 6, 2010
0
10
20
30
July 4, 2010
0
4
8 July 4, 2010
0
20
40
60
January 7, 2011
0
4
8 January 7, 2011
0
10
20
30
40
May 3, 2011
0
4
8
12
May 3, 2011
0
4
8
12
April 7, 2012
0
10
20
30
40
May 14, 2013
0
4
8
12May 14, 2013
50 55 60 65 70
0
4
8
12March 1, 2014
Radial velocity, km s-1
Fig. 1. Spectra of OH maser emission in the 1665- and 1667-MHz linesat various epochs. Solid curves: left-hand circular polarisation; dashed:right-hand circular.
Figures 2 and 3 show Stokes parameters for the main lines at1665 and 1667 MHz at the epochs of January 6, 2008, and May3, 2011. Figure 4 presents central parts of the spectra for StokesparameterV of the OH main lines at different epochs. The mainspectral features are numbered. The results of our observationsin the 1612- and 1720-MHz satellite lines are presented in Fig-ures 5 and 6, respectively.
Figure 7 represents an atlas of the H2O spectra for the inter-val from November 1981 to January 2014. For technical reasons,no observations were conducted between May 2006 and Decem-ber 2007. The horizontal axis is the velocity relative to thelocalstandard of rest (LSR). All the spectra are given in the sameradial-velocity scale. An arrow at the vertical axis shows thescale in janskys. In the spectra zero baselines have been drawn.
Superpositions of the H2O spectra for various time intervals(1–6) are shown in Figure 8. The separation was done accord-ing to the character of the spectral evolution. Averaged spec-tra are shown with bold curves. The maser emission in the ve-locity interval of 55 – 63 km s−1 is observed during two timeintervals (1 and 3). At the rest time the emission is observedmainly in one or two narrow velocity intervals, 55 – 57.5 and58 – 62 km s−1. From 2001 to 2012 the velocity centroid of theaveraged spectra of the main group (58 – 62 km s−1) moved from
Article number, page 2 of 18
P. Colom et al.: OH and H2O maser variations in W33B
0
10
20
30
I
January 6, 2008
1665 MHz
Flu
x de
nsity
, Jy
-1
0
1
2
Q
-2
-1
0
1 U
50 55 60 65 70
-20
-10
0
10V
Radial velocity, km s-1
0
5
10
15
I1667 MHz
-0.8
-0.4
0.0
0.4
0.8Q
-0.8
-0.4
0.0
0.4
0.8U
50 55 60 65 70-8
-4
0
4V
Fig. 2. Stokes parameters of the main lines at 1665 and 1667 MHz forthe epoch January 6, 2008.
60.4 to 59.4 km s−1. In addition, fluctuations of the velocity cen-troid calculated for individual spectra were observed.
0
20
40
60
I
May 3, 2011
1665 MHz
Flu
x de
nsity
, Jy
-1
0
1
2
3
Q
-2
-1
0
1 U
50 55 60 65 70-40
-30
-20
-10
0
10V
Radial velocity, km s-1
0
5
10
15
20
I1667 MHz
-1.0
-0.5
0.0
0.5
1.0Q
-1.0
-0.5
0.0
0.5
1.0U
50 55 60 65 70-15
-10
-5
0
5V
Fig. 3. Same as in Fig. 2, for the epoch May 3, 2011.
56 60 64 68
March 1, 20140
0
0
May 14, 2013
6543
May 3, 2011
January 7, 2011
July 4, 2010
April 6, 201010 Jy
January 6, 20081667 MHzV
March 8, 2012
April 7, 2012
56 60 64 68
March 1, 2014
May 14, 20130
0
0
0
0
0
21
0
0
0
0
0
May 3, 2011
January 7, 2011
July 4, 2010
April 6, 2010
January 6, 2008
30 Jy
V
1665 MHz
Flu
x de
nsity
, Jy
Radial velocity, km s-1
Fig. 4. Central parts of the OH main-line spectra for the Stokes parame-ter V for various epochs of the observations. Main spectral features arenumbered.
3. Data analysis and discussion
3.1. Hydroxyl
In 1978, using high angular resolution in right ascension oftheNançay radio telescope, we searched for the emission/absorptionpeaks in the satellite OH lines. Figure 9 shows the emi-sion/absorption intensity as a function of right ascension for dif-ferent spectral features. Vertical arrows denote positions of radiocontinuum peaks (W33C and B). Features 1 (62.3 km s−1) and2 (56 km s−1) are unresolved, whereas feature 3 (∼50 km s−1) isextended. Peaks of features 1 and 2 in the 1612-MHz line do notcoincide; the former is detectable only in the 1612-MHz lineandcan be associated with a maser spot.
Thus, according to our observations in 1978, 1991, 2012, and2014, (thermal) emission/absorption satellite-line features in thevelocity interval 49–58 km s−1 belong to an extended source,whereas narrow emission features come from a pointlike source.
Of interest is broadband absorption in the main OH lines,which probably covers the entire velocity interval filled byOHmaser emission features. It is visible at theI profile edges and,probably, in theU Stokes profile (Figures 2 and 3), thus sug-gesting a slight linear polarisation. The velocity range can beassessed more certainly in the OH satellite lines, especially in1720 MHz, where OH is partly in thermal emission and partlyin absorption (see Figures 5 and 6). It covers radial velocitiesVLSR ≈ 50–65 km s−1, which probably represent the entire ve-locity dispersion of material in the molecular cloud surroundingthe W33B maser.
Article number, page 3 of 18
A&A proofs:manuscript no. aa23083-13
48 52 56 60 64
March 1, 2014
1 Jy
LR
March 8, 2012
May 10, 1991
March 28, 1981
October 20, 19781612 MHz
Flu
x de
nsity
, Jy
Radial velocity, km s-1
Fig. 5. Spectra of OH maser emission in the satellite line at 1612 MHz.Spectra in right- and left-hand circular polarisation are shown with dot-ted and solid lines, respectively. Strong narrowband polarised maseremission is visible in the 2012 and 2014 profiles.
3.1.1. The structure of the OH maser source
Figure 10a shows the arrangement of the main-line OH masersspots for the epoch 1991 (Argon et al. 2000). Spots’ radial ve-locities are indicated. The angular size of the main-line maseringregion is 0.4×0.4 arcsec; the regions of emission in the 1665-and 1667-MHz lines are spatially separated. The map centre∆RA = 0, ∆Dec = 0 corresponds to RA(2000)=18h13m54.75s,Dec(2000)=−18◦1′46.4′′. The distribution of the maser spots isdelineated with arcs (dashed curves). We observe regular radial-velocity variations along the arc for the 1665-MHz maser spots.At first, the velocity decreases, then increases. There are severalclusters of maser spots with a smallVLSR dispersion. An asteriskmarks the presumed location of the central star for the OH masersource based on the arc shape i.e. on the large-scale structure ofthe source. Another argument supporting our supposition isthelarge separation between the OH and H2O masers in W33B. Incontrast to Forster & Caswell (1989), we think that each maserhas its own source of energy.
45 50 55 60 65 70
May 14, 2013
L
R
2 Jy
L
R
May 10, 1991
April 7, 2012
Linear
L
R
March 28, 1981
1720 MHz
October 21, 1978
R
L
Flux
den
siry
, Jy
Radial velocity, km s-1
Fig. 6. Spectra of OH maser emission in the satellite line at 1720 MHz.There is no difference between right- (R) and left-hand (L) polarisedprofiles.
3.1.2. Main-line emission
The OH maser in W33 B is a typical representative of hydroxylmasers associated with star-forming regions: the 1665-MHzlineis stronger than the 1667-MHz line.
Our main-line observations have shown that most OH emis-sion features in W33 B have a high degree of circular polari-sation (see Figure 1). Another peculiarity of the OH spectrais
Article number, page 4 of 18
P. Colom et al.: OH and H2O maser variations in W33B
strong variability of the most intense features. We have found noZeeman splitting for such features.
For the detection and study of Zeeman components we useStokes parameterV, the difference of the right- and left-handcircular polarisations (see Figures 2–4). In the 1665-MHz linetheσ Zeeman components can be a pair of spectral features near62 km s−1, and in the 1667-MHz line these are pairs of featuresnear 62 and 64 km s−1. The splitting is 1.3, 0.7 and 1.1 km s−1.This corresponds to intensities of the line-of-sight magnetic fieldof B = −2.2,−2.0 and−3.1 mG, respectively (B < 0 correspondsto the field directed toward the observer).
It should be emphasised that in both OH main lines at62 km s−1 we have obtained quite similar field intensities,−2.2and −2.0 mG. These features may be produced by the samemaser condensation.
As for previous measurements of the magnetic field towardW33 B, Zheng (1991), who observed W33B on the 43-m NRAOradio telescope in October 1981, found an average fieldB =−5 mG from the shifts of the mean weighted velocities in theboth main OH lines 1665 and 1667 MHz. Fish et al. (2003) foundin their VLA observations of August 1991 several Zeeman pairsof right- and left-hand circularly polarised features within a ve-locity interval of 62.5–64.5 km s−1 in both OH main lines. Theirestimates for the field strength are from−0.7 to−7.5 mG. Thequoted estimates do not contradict ours, both in the field direc-tion and in the order of magnitude. The direction of the mag-netic field fits the general field pattern in the Scutum Arm asdirected counterclockwise when viewed from the north galacticpole. Thus, during the collapse of the protostellar gas cloud thegeneral direction of the interstellar field could have been con-served.
3.1.3. Satellite-line emission
As in the source W33C in the same region (Colom et al. 2012),the profiles of the 1612- and 1720-MHz OH lines toward W33B(Figures 5, 6) consist of emission and absorption componentsand mirror each other: peak velocities are, respectively, 56.15and 51.66 km s−1 in the 1612-MHz line, 51.47 and 56.23 km/sin the 1720-MHz line. The mean difference is 4.6 km s−1 (forW33C it is 3.4 km s−1).
This structure is explained in a model of an OH source asso-ciated with a molecular cloud where a maser is embedded in. Ifan IR source is present in the cloud its radiation affects the pop-ulations of the hyperfine structure sublevels of OH molecules(Burdyuzha & Varshalovich 1973). In this model particularsofthe IR pumping are such that inversion of the 1720-MHz tran-sition levels is accompanied by anti-inversion in the 1612-MHztransition and vice versa. The inversion or anti-inversionis de-termined by the angle between the direction of propagation ofthe IR radiation and the direction of local magnetic field. For asource embedded in the cloud one of the satellites is inverted andthe other one is anti-inverted; in the other part of the cloudthesituation is the opposite. This effect was observed by us in W33C(Figure 9).
In 2012 we detected narrowband 1612-MHz maser emis-sion in both circular polarisations. In 1991 this emission wasabsent. We detected three emission features at radial velocities59.1, 59.7 and 60.4 km s−1. Table 1 lists their flux densitiesat different epochs. The columns labelledFR and FL containflux densities in the right- and left-hand circular polarisations,respectively; the columns labelledp list degree of polarisationp =| FR − FL | / (FR + FL) .
50 55 60 65
400 Jy
May 6, 1982
June 17, 1981
(1)W33B
February 2, 1982
June 10, 1982
December 16, 1981
October 28, 1981
February 17, 1981
February 11, 1981
March 18, 1982
Flu
x de
nsity
, Jy
Radial velocity, km s-150 55 60 65
October 5, 1983
October 13, 1982
December 30, 1982
March 17, 1983
May 5, 1983
June 8, 1983
400 Jy
November 14, 1984
December 16, 1983
(2)
June 12, 1984
December 11, 1984
April 4, 1984
February 9, 1984
September 25, 1984
Fig. 7. Spectra of the H2O maser emission in W33B.
Table 1. Flux densities in the right- (FR) and left-hand (FL )circular polarisations in janskys and the degree of polarisation pin the 1612-MHz OH line.Typical 1σ errors are about 0.1 Jy.
Radial velocity, km s−1
Date 59.1 59.8 60.4FR FL p FR FL p FR FL p
2012 Mar 8 1.4 0.3 0.65 2.7 2.6 0.02 1.0 6.9 0.752012 Apr 7 1.6 0 1 3.7 2.5 0.19 0.3 7.5 0.922012 Sept 7 1.7 0 1 4.5 2.8 0.23 0.6 9.4 0.88
2013 May 14 1.5 0 1 5.5 3.4 0.24 0.7 10.6 0.882014 Mar 1 1.5 0 1 5.1 2.6 0.32 0.3 10.7 0.95
The emission at 59.8 and 60.4 km s−1 is variable, and thedegree of polarisation also varies. The cause of this can be tur-bulent motions of material in the masering region. The observedemission components have no strong counterparts either in themain OH lines or in the 1720-MHz satellite line. Thus, we mayobserve a new Type IIb OH maser probably associated with aninfrared star. However, the radial-velocity range (∼1.3 km s−1)is too narrow as compared to typical values for OH masers inlate-type giants and supergiants, from 5 km s−1 up to 40 km s−1.The 1612-MHz emission features being within the velocity in-terval of the main-line emission suggests the same distanceandphysical association with the 1665/1667-MHz maser. Very longbaseline interferometry (VLBI) observations would be desirableto resolve this question.
Article number, page 5 of 18
A&A proofs:manuscript no. aa23083-13
0
200
400 XII.1992 - VIII.1997
(2)
52 56 60 64-100
0
100
200
300
400
500
X.1998 - V.2001(3)
Radial velocity, km s-1
0
400
800
1200
1600
2000
240015800 Jy
(5)III.2005 - VII.2008
0
400
800
1200
1600
2000
2400
(4)
IX.2001 - XI.2004
0
200
400
600 (1)
II.1981 - XI.1988
Flu
x de
nsity
, Jy
52 56 60 64
0
200
400
600
800
1000
1200
(6)IV.2009 - V.2012
Fig. 8. Superposition of the H2O 22-GHz spectra for various time inter-vals. The separation was done according to the character of the spectralevolution. Averaged spectra for the intervals are shown with bold lightcurves. See text for details.
3.2. Water-vapour emission
On the basis of our regular long-term monitoring of W33B per-formed with a high spectral resolution, we have studied bothfastand long-term variations of H2O maser emission as well as theevolution of individual features.
Figure 11 shows variations of the integrated H2O line flux.The point marked (×) is from Lada et al. (1981) and (△) is fromJaffe et al. 1981. Two stages of maser activity are prominent.Arrows mark the epochs of the interferometric observations.
We find no cyclicity in the maser variability. We observedtwo deep minima: in 1989–1990 and in mid-1997. In addition,we observed flare-type variability.
There were several interferometric observations of the H2Omaser in W33B. In January 1979 W33B was observed for thefirst time by VLBI in the 22-GHz H2O line by Lada et al. (1981)with a baseline of 845 km. Their map contains four maser fea-tures within a region∼ 0.05′′ i.e. 120–250 AU depending on theaccepted distance, 2.4 or∼5 kpc (see Introduction).
In June 1984, Forster & Caswell (1989) observed W33B onVLA with a spatial resolution of about 3′′ and radial veloc-ity resolution of 1.32 km s−1. The estimated size of the regionhosting maser spots (except for one) is 0.06′′ × 0.07′′ (300×350AU) in the right ascension and declination, respectively (see Fig-
30 20 10 0 -10 -20 -30
-3
-2
-1
0
1
2
-2
-1
0
1
2
3
4b
5
3b
2
50 km s-1
56 km s-1
36 km s-1
32 km s-1
1720 MHz
RAS
5
4a1612 MHz
32 km s-1
35 km s-1
51 km s-1
56 km s-1
62.3 km s-1
3a
2
1
W33 C W33 B
Flux
den
sity
, Jy
Fig. 9. Flux density of the features in the satellite OH lines as a functionof right ascension toward W33.∆RA=0 corresponds to the position ofW33C. Features with similar radial velocities are designated3a, 4a and3b, 4b in the 1612- and 1720-MHz lines, respectively. The strong 1612-MHz maser feature at 62.3 km s−1 (shown with bold line in the upperpanel) certainly belongs to W33B.
ure 9b). The map centre∆RA = 0, ∆Dec = 0 corresponds toRA(2000)=18h13m54.7s, Dec(2000)=−18◦1′48.0′′. The maserspots observed in 1979 are plotted with filled circles, and thoseobserved in 1984 with open circles. The radial velocities ofeachmaser spot are indicated. The presence of a radial-velocitygra-dient is visible, its direction is shown with a dashed line. Ahor-izontal bar shows the linear scale of the W33B region.
Immer et al. (2013) observed H2O masers in the W33 com-plex on VLBA at nine epochs between September 2010 and Jan-uary 2012. In addition to the trigonometric parallax yielding adistance of 2.4 kpc, they found proper motions of individualH2Omaser features in W33B that are within 4 mas/year. We note thatsuch proper motions correspond to projected velocities in thesky plane of∼45 km s−1, which are by a factor of a few largerthan the maximum line-of-sight velocity spread (∼10 km s−1, seeFigure 8). This is at variance with the model of spherically sym-metric expansion; the cause of this difference is not yet clear.
We have identified main spectral components of our monitor-ing (for the period June 1984) with the VLA map components.This is shown with arrows in the left panel of Figure 12. Sincethe spectral resolution of our monitoring is considerably higher,
Article number, page 6 of 18
P. Colom et al.: OH and H2O maser variations in W33B
0.1 0.0 -0.1 -0.2 -0.3
-0.2
-0.1
0.0
0.1
0.2
0.10 0.05 0.00 -0.05
-0.05
0.00
0.05
64.46
64.95
65.2461.35
63.8659.11
63.4664.04
61.16
65.5365.32
62.93
59.6862.37
(a)
100 AU
OH
LR
LR
1667 MHz
1665 MHz
Dec
"
60.58
60.78
66.08
66.20
65.6463.72
65.66
59.3854.92
61.6656.89
61.0964.29
60.01
64.8164.5
63.6463.37
62.5262.56
62.7562.14
64.42
59.0450.7458.57
63.3461.73
62.28
22 GHz
(b)
RA"
100 AU
H2O
54.7362.64
53.41
56.05
57.36
58.6860.00
61.32
Dec
"
RA"
Fig. 10. (a) VLA map of main-line OH masers for the epoch 1991 (Ar-gon et al. 2000). Right- and left-hand polarised features are shown withdifferent symbols. Spots’ radial velocities are indicated, those of left-hand polarised features are initalics. An asterisk marks the presumedlocation of the central star for the OH maser source based on the arcshape.(b) Positions of VLA H2O maser spots in June 1984 (Forster & Caswell1999) On both maps the 100 AU bar corresponds to the trigonometricdistance of 2.4 kpc (Immer et al. 2013). Spots’ radial velocities are in-dicated. Dotted curves show the preferential direction of the growth ofthe spots’ radial velocities.
we have identified and traced the evolution of a larger numberofH2O emission features than the VLA maps. Thus, each featureon the VLA map may actually correspond to a cluster of severalmaser spots. TheVLSR gradient testifies that the clusters of maserspots form a large-scale organised structure.
Since 2000 the H2O maser was quite active in the radial-velocity interval 58–62 km s−1, with an intermediate minimumin early 2005. Since mid-2011 the activity has declined. In thequoted activity interval we observed flares of individual featuresas well as of groups of features. The strongest flare took place in2006. In March the flux density reached almost 16,000 Jy.
Figure 12 (right-hand panel) shows radial-velocity variationsof the strongest emission features in W33B. For each of themcircles of different sizes mark the emission maxima in whichthe flux density exceeded 700 Jy. The circle size correspondsto the flux density magnitude. For all maxima flux densities
1980 1985 1990 1995 2000 2005 2010 2015
0
1000
2000
3000
4000
5000
600012600 Jy
Inte
grat
ed f
lux,
Jy
km s
-1
Years
Fig. 11. Variations of the integrated H2O line flux. Two points at the lefttaken from observations of other authors: (×) from Lada et al. (1981)and (△) from Jaffe et al. 1981. Arrows mark the epochs of the interfero-metric observations.
are given in janskys. In 2000–2003 the emission maximum wasmoving toward higher velocities and then in the opposite direc-tion (dashed curves). This may be due to an ordered arrangementof maser condensations such as clusters, fragments of shells, etc.This result is consistent with the H2O structure on the VLA map(Forster & Caswell 1999) as well as with the map of Immer et al.(2013).
The strongest flare at 59.4 km s−1 in the beginning of 2006was preceded by complicated variations in the spectrum struc-ture and by a shift of the emission peak from 60.6 to 59.4 km s−1.At the epoch of the maximum the line profile was gaussian. Theline was symmetric, its width at half-maximum was 0.53 km s−1.It was local, and it might be associated with a maser spot. In thepresence of turbulent or chaotic motions of maser condensationsin a cluster, two clumps of material can become superposed inthe line of sight, which results in an increase in the opticaldepthτ of the medium. For instance, in the case of the unsaturatedmaser an increase inτ by a factor of 2.3 results in an intensitygrowth by an order of magnitude.
In some time intervals we observed appreciable radial-velocity variations of the H2O emission features. Most likely,each feature is identified not with an individual maser spot butwith some structure, for instance, a filament, a chain with aradial-velocity gradient or with a more complex formation;seee.g. Torrelles et al. (2003), Lekht et al. (2007). We observed theemission from such structures during intervals of high activityof the maser source, often in the periods of flare activity. Duringpropagation in the masering medium of a shock wave driven bythe stellar wind or molecular outflow, regions with different ra-dial velocities are consecutively excited. This results inthe spec-tral and spatial drift of the observed emission peak. In addition,we observed an appreciable radial-velocity drift of the emissionfor the cluster of maser spots from 57.5 to 55.5 km s−1 (dashedline in Figure 12b). The drift is indeed considerable to takeintoaccount that the radial-velocity dispersion of all maser spots inW33B does not exceed 8 km s−1. The drift time interval inter-val falls on the stage of maximum H2O maser activity in W33B.This is confirmed by the proper motions of the maser condensa-tions found by Immer et al. (2013).
The complicated character of the radial-velocity variationsmay reflect the presence of turbulent motions of material within
Article number, page 7 of 18
A&A proofs:manuscript no. aa23083-13
1998 2000 2002 2004 2006 2008 2010 2012 2014
440
(b)
480
500
260
2200 < F1400 < F < 22001000 < F < 1400700 < F < 1000
14252160
805
727
920
1650
1265
795 1160
1220
11351030
1050
213015770
Years1982 1984
55
56
57
58
59
60
61
62
63
(a)
Rad
ial v
eloc
ity, k
m s
-1
Fig. 12. Radial-velocity variations of the strongest H2O emission features in W33B. Peaks at different velocities are shown with different symbols.Main features are shown with circles of different kinds. Circles of various sizes denote emission peakswith flux densities that exceeded 700 Jy.The circle size corresponds to the flux density magnitude. For each peak its flux density is given in janskys. Dashed and dotted lines (drawn byeye) present probable radial-velocity drifts of features persisting throughout our monitoring. Characteristic timescales are of the order of months.
an H2O maser condensation as well as on the scale of compactclusters of maser spots.
4. Summary
We list the main results of our long-term observations of thehy-droxyl maser and of the 30-year monitoring of the water-vapourmaser in the source W33 B.
1. We have observed strong variability of the emission featuresin the main OH lines 1665 and 1667 MHz.
2. We have detected Zeeman splittingσ-components in the1665-MHz OH line at 62 km s−1 and in the 1667-MHz lineat 62 and 64 km s−1. From the amount of splitting we haveestimated the line-of-sight component of the magnetic fieldfor each of the masering regions as 2.0–2.2 and 3.1 mG, re-spectively.
3. The profiles of the satellite lines at 1612 and 1720 MHz mir-ror each other. This suggests pumping of the levels of thesetransitions by infrared emission of the source embedded inthe magnetised interstellar cloud around the maser.
4. We have detected narrowband, strongly variable emissioninthe 1612-MHz line with a high degree of circular polarisa-tion, which belongs to a pointlike source.
5. We present an atlas of the H2O λ = 1.35 cm emission spec-tra toward W33 B for the time interval from November 1981to May 2006 and from December 2007 to January 2014.The mean interval between consecutive observational ses-sions was 2.2 months. The radial-velocity resolution was0.101 km s−1, and since the end of 2005 it was 0.0822 km s−1.
6. We detected flares of individual H2O spectral features andof groups of features (clusters). The emission features prob-ably form filaments, chains with a radial-velocity gradients,or more complicated structures including large-scale ones.
7. The characteristic variations of OH and H2O maser emissionsuggest the existence of turbulent motions of material in theregions of the maser spots’ localisation.
8. We have observed two stages of activity of the H2O maserwith an interval between the main maxima of about 20 years.
9. The arc-like arrangement of the OH maser spots and the largeseparation between the OH and H2O maser sources allow usto suppose that the hydroxyl and water vapour masers haveindependent energy sources.
Acknowledgements. The Nançay Radio Observatory is the Unité Scientifique deNançay of the Observatoire de Paris, associated with the CNRS. The NançayObservatory acknowledges the financial support of the Région Centre in France.The 22 m Pushchino radio telescope is supported by the Ministry of Science
Article number, page 8 of 18
P. Colom et al.: OH and H2O maser variations in W33B
and Education of the Russian Federation (facility registration number 01-10).This work was supported by the Russian Foundation for Basic Research (projectcode 09-02-00963-a). The authors are grateful to the staff of the Nançay andPushchino Radio Astronomy Observatories for their help with the observationsand to the anonymous referee for useful comments that helpedto improve thepaper. This research has made use of the SIMBAD database, operated at CDS,Strasbourg, France
ReferencesArgon, A. L., Reid, M. J., & Menten, K. M. 2000, ApJS, 129, 159Burdyuzha, V. V. & Varshalovich, D. A. 1973, Soviet Astron.,16, 597Caswell, J. L., Green, J. A., & Phillips, C. J. 2013, MNRAS, 431, 1180Colom, P., Lekht, E. E., Pashchenko, M. I., & Rudnitskii, G. M. 2012, Astron.
Rep., 56, 731Fish, V. L., Reid, M. J., Argon, A. L., & Menten, K. M. 2003, ApJ, 596, 328Forster, J. R. & Caswell, J. L. 1989, A&A, 213, 339Forster, J. R. & Caswell, J. L. 1999, A&AS, 137, 43Gardner, F. F., Wilson, T. L., & Thomasson, P. 1975, Astrophys. Lett., 16, 29Genzel, R. & Downes, D. 1977, A&AS, 30, 145Goss, W. M. 1968, ApJS, 15, 131Goss, W. M., Matthews, H. E., & Winnberg, A. 1978, A&A, 65, 307Goss, W. M. & Shaver, P. A. 1970, Austral. J. Phys. Astrophys.Suppl., 14, 1Haschick, A. D. & Ho, P. T. P. 1983, ApJ, 267, 638Immer, K., Reid, M. J., Menten, K. M., Brunthaler, A., & Dame,T. M. 2013,
A&A, 553, A117Jaffe, D. T., Güsten, R., & Downes, D. 1981, ApJ, 250, 621Kolpak, M. A., Jackson, J. M., Bania, T. M., Clemens, D. P., & Dickey, J. M.
2003, ApJ, 582, 756Lada, C. J., Blitz, L., Reid, M. J., & Moran, J. M. 1981, ApJ, 243, 769Lekht, E. E., Pashchenko, M. I., & Rudnitskii, G. M. 2012, Astron. Rep., 56, 45Lekht, E. E., Slysh, V. I., & Krasnov, V. V. 2007, Astron. Rep., 51, 967Menten, K. M. 1991, ApJ, 380, L75Pandian, J. D., Momjian, E., & Goldsmith, P. F. 2008, A&A, 486, 191Pashchenko, M. I. 1980, Soviet Astron. Lett., 6, 58Pashchenko, M. I., Rudnitskiı̆, G. M., & Colom, P. 2009, Astron. Rep., 53, 541Quireza, C., Rood, R. T., Balser, D. S., & Bania, T. M. 2006, ApJS, 165, 338Robinson, B. J., Goss, W. M., & Manchester, R. N. 1970, Austral. J. Phys., 23,
363Sato, M., Wu, Y. W., Immer, K., et al. 2014, ApJ, 793, 72Slysh, V. I., Pashchenko, M. I., Rudnitskiı̆, G. M., Vitrishchak, V. M., & Colom,
P. 2010, Astron. Rep., 54, 599Torrelles, J. M., Patel, N. A., Anglada, G., et al. 2003, ApJ,598, L115Zheng, X.-W. 1991, Chin. J. Space Sci., 11, 1
List of Objects
‘W33’ on page 1‘G12.80−0.20’ on page 1‘G12.68−0.18’ on page 1‘G12.80−0.18’ on page 1‘W33’ on page 1‘G12.68−0.18’ on page 1‘G12.80−0.20’ on page 1‘G12.91−0.28’ on page 1‘W33B’ on page 5
Article number, page 9 of 18
A&A–aa23083-13,Online Material p 10
50 55 60 65
400 Jy
May 6, 1982
June 17, 1981
(1)W33B
February 2, 1982
June 10, 1982
December 16, 1981
October 28, 1981
February 17, 1981
February 11, 1981
March 18, 1982
Flu
x de
nsity
, Jy
Radial velocity, km s-150 55 60 65
October 5, 1983
October 13, 1982
December 30, 1982
March 17, 1983
May 5, 1983
June 8, 1983
400 Jy
November 14, 1984
December 16, 1983
(2)
June 12, 1984
December 11, 1984
April 4, 1984
February 9, 1984
September 25, 1984
Fig. 7. Spectra of the H2O maser emission in W33B.
A&A–aa23083-13,Online Material p 11
50 55 60 65
1986
1985
1985
March 29, 1988
February 17, 1988
January 7, 1988
December 3, 1987
200 Jy
February 24, 1987
November 11,
(3)W33B
September 17, 1986
May 29, 1987
January 28, 1986
December 12,
October 3, 1985
June 26, 1985
November 1,
Flu
x de
nsity
, Jy
Radial velocity, km s-150 55 60 65
June 24, 1993
April 22, 1993
December 16, 1992
March 2, 1993
April 15, 1992
200 Jy
July 4, 1990
May 4, 1988(4)
December 26, 1989
January 17, 1992
March 28, 1991
November 11, 1988
September 21, 1988
March 27, 1990
Fig. 7. Continued.
A&A–aa23083-13,Online Material p 12
50 55 60 65
1997September 16,
August 20, 1997
May 7, 1997
July 21, 1997
100 Jy
February 25, 1997
March 20, 1996
(5)W33B
December 10, 1996
April 1, 1997
October 7, 1996
July 5, 1996
July 4, 1995
February 9, 1995
January 21, 1997
Flu
x de
nsity
, Jy
Radial velocity, km s-150 55 60 65
May 12, 1999
April 1, 1999
100 Jy
October 6, 1998
January 23, 1998(6)
July 1, 1998
January 19, 1999
December 30, 1998
June 2, 1998
February 26, 1998
July 21, 1998
Fig. 7. Continued.
A&A–aa23083-13,Online Material p 13
50 55 60 65
2000
1999
1999
200 Jy
April 26, 2000
November 1, 1999
(7)W33B
February 22, 2000
January 26,
December 15,
September 28,
June 16, 1999
March 31, 2000
Flu
x de
nsity
, Jy
Radial velocity, km s-150 55 60 65
September 4, 2001
200 Jy
March 27, 2001
May 30, 2000
(8)
May 29, 2001
April 12, 2001
June 14, 2000
August 29, 2000
January 17, 2001
Fig. 7. Continued.
A&A–aa23083-13,Online Material p 14
50 55 60 65
September 13, 2002
August 12, 2002
July 23, 2002
June 5, 2002
400 Jy
December 26, 2001
(9)W33B
April 23, 2002
February 5, 2002
March 28, 2002
November 28, 2001
October 23, 2001
Flu
x de
nsity
, Jy
Radial velocity, km s-150 55 60 65
400 Jy
January 28, 2003
March 26, 2003
(10)
October 10, 2002
April 23, 2003
November 11, 2002
Fig. 7. Continued.
A&A–aa23083-13,Online Material p 15
50 55 60 65 50 55 60 65
2003
1590 Jy
400 Jy
September 24, 2003
(11)W33B
December 1,
October 27, 2003
January 28, 2004
July 4, 2003
May 28, 2003
Flu
x d
ensi
ty, Jy
Radial velocity, km s-1
2004
2004
March 11, 2004
March 15, 2005
February 1, 2005
November 4,
September 27,
400 Jy
September 1, 2004
April 21, 2004
(12)
June 16, 2004
May 26, 2004
July 20, 2004
Fig. 7. Continued.
A&A–aa23083-13,Online Material p 16
50 55 60 65
January 30, 2006
December 14, 2005
400 Jy
August 22, 2005
(13)W33B
December 21, 2005
September 26, 2005
November 9, 2005
June 28, 2005
April 14, 2005
Flu
x de
nsity
, Jy
Radial velocity, km s-150 55 60 65
July 2, 2008
2000 Jy
June 18, 2008
March 28, 2006
(14)
May 14, 2008
April 9, 2008
April 26, 2006
Fig. 7. Continued.
A&A–aa23083-13,Online Material p 17
50 55 60 65
April 27, 2010
2009
2009
March 30, 2010
February 26, 2010
December 14, 2009
400 Jy
August 24,
(15)W33B
January 28, 2010
October 6, 2009
November 9,
June 16, 2009
April 15, 2009
Flu
x de
nsity
, Jy
Radial velocity, km s-150 55 60 65
May 29, 2012
April 24, 2012
March 28, 2012
February 28, 2012
January 31, 2012
940 Jy
December 14, 2011
November 24, 2011
October 24, 2011
September 2, 2011
July 27, 2011
June 30, 2011
June 15, 2011
November 3, 2010
May 26, 2011
400 Jy
March 28, 2011
(16)
February 7, 2011
March 3, 2011
October 10, 2010
Fig. 7. Continued.
A&A–aa23083-13,Online Material p 18
50 55 60 65
January 28, 2014
November 25, 2013
August 16, 2013
September 16, 2013
October 28, 2013
July 18, 2013
February 27, 2013
November 22, 2012
400 Jy
July 30, 2012
(17)W33B
January 28, 2013
September 24, 2012
October 29, 2012
August 29, 2012
July 3, 2012
Flux
den
sity
, Jy
Radial velocity, km s-1
Fig. 7. Continued.