long-term monitoring of the w44c (g 34.3+0.15) maser in the 1.35 cm water vapor and 18 cm hydroxyl...

ISSN 1063-7729, Astronomy Reports, 2012, Vol. 56, No. 1, pp. 45–62. c© Pleiades Publishing, Ltd., 2012.Original Russian Text c© E.E. Lekht, M.I. Pashchenko, G.M. Rudnitskii, 2012, published in Astronomicheskii Zhurnal, 2012, Vol. 89, No. 1, pp. 48–65.

Long-Term Monitoring of the W44C (G 34.3+++0.15) Maserin the 1.35 cm Water Vapor and 18 cm Hydroxyl Lines

E. E. Lekht*, M. I. Pashchenko, and G. M. RudnitskiiMoscow State University, Sternberg Astronomical Institute,

Universitetskii pr. 13, Moscow, 119234 RussiaReceived April 21, 2011; in final form, August 17, 2011

Abstract—Results of monitoring the H2O and OH masers in W44C, located near the cometary HII regionG34.3+0.15, are reported. Observations in the water-vapor line at λ = 1.35 cm were carried out on the22-meter radio telescope of the Pushchino Radio Astronomy Observatory (Russia) from November 1979to March 2011, and in the hydroxyl lines at λ = 18 cm on the large Nancay radio telescope (France).Activity maxima and minima of the water maser alternated. The average period of the activity is ∼ 14 years,consistent with results obtained earlier for a number of other sources associated with regions of active starformation. In periods of enhanced maser activity, two series of strong H2O maser flares were observed,which were related to two different clusters of maser spots located at the front of a shock at the peripheryof the ultracompact region UH II. These series of flares may be associated with cyclic activity of theprotostellar object in UH II. In the remaining time intervals, there were mainly short-lived flares of singlefeatures. The Stokes parameters for the observations in the hydroxyl lines were determined. Zeemansplitting was observed in the profile of the 1667 MHz OH main line at a velocity of 58.5 km/s, and wasused to estimate the intensity of the line-of-sight component of the magnetic field (1.2 mG).

DOI: 10.1134/S1063772912010040

1. INTRODUCTION

The source G34.3+0.15 is in a region of active starformation (distance 3.8 kpc), which hosts three com-pact and one extended HII region [1] (Fig. 1). One ofthese (component C) is an ultracompact HII regionwith a cometary morphology, which is embedded inan ultracompact molecular core with a temperatureof about 225 K and a hydrogen density of 107 cm−3

[2]. The maser emission of G34.3+0.15 (W44C) wasdiscovered by Turner and Rubin [3], first in the mainhydroxyl lines, 1665 and 1667 MHz, and then in the22 GHz water line in 1971.

According to the VLA observations of Fey et al.[4], there are three separate groups of H2O maserspots in a 50′′ × 30′′ area in the G34.3+0.15 complex.The largest number of maser spots coincides withthe hot core of the molecular cloud, i.e., they aredirectly connected to the cometary-type HII region(component C). Two other groups of maser spots areseparated from the hot core by 2′′ and 4′′; these arenot associated with any continuum source. Theirvelocities do not differ strongly from those of the spotsin the hot core. The H2O maser spots associated withcomponent C (the cometary HII region) form severalclusters [5]. The three main clusters are extended in

*E-mail:[email protected]

the radial direction relative to the center of the HIIregion.

Subsequent interferometric observations (see,e.g., [6–9]) showed that the H2O and OH masers arearranged along a parabolic arc near the front of a bowshock at the eastern edge of the cometary HII region.The OH masers are closer to the edge of the cometaryregion than the H2O masers, suggesting that the OHmolecules are formed via the dissociation of H2Omolecules in the shock; the OH masers probablytrace the position of a strong shock that separatesthe cool molecular cloud from a dense warm envelopeencompassing the bow shock [9].

According to the 13-year (1987–1999) observa-tions of Valdettaro et al. [10], there was fairly stableH2O emission at a radial velocity of about 60 km/s,together with short-lived strong flares at other veloc-ities. In addition, the integrated flux varied by at mosta factor of three.

The observations in molecular lines, e.g., C34S[11], 13CO, CS, and CH3CN [12], revealed in thisregion the presence of a strong North–South radial-velocity gradient on a scale of ∼ 40′′. This testifies tothe existence of a flow of molecular material in thisdirection. The velocity gradient was also detectedfrom observations in the H93α recombination line[13].

45

46 LEKHT et al.

45.845.945.046.146.246.346.418h50m46.5s

1°11′17″

8

16

15

14

13

12

11

10

9

Dec

lin

atio

n (

1950

.0)

Right ascension

Shockfront

A

10

5

B

3

2

98

4

76

C

1

Fig. 1. Schematic representation of the G34.3+0.15 region. The crosses show the positions of clusters of H2O masers (thelarge crosses show the main H2O clusters), and the circles the positions of the main OH emission regions.

We first observed W44C in all four 18-cm linesof the OH molecule on the large radio telescope inNancay (France) in 1974 [14]. W44C was also ob-served in OH lines on the Nancay radio telescopeat various later epochs. Since 1979, regular obser-vations (monitoring) of W44C in the H2O line hasbeen carried out on the 22-m radio telescope in thePushchino Radio Astronomy Observatory [15].

2. OBSERVATIONS AND DATAThe observations of the H2O maser emission in

the 1.35-cm line toward G34.3+0.15 were carriedout on the 22-m radio telescope in Pushchino fromNovember 1979 to March 2011. The mean intervalbetween observations was less than two months. Thenoise temperature of the system, which had a cooledFET front-end amplifier, was 150–250 K. An up-grade of the receiver in 2000 lowered the system noisetemperature to 100–150 K, depending on weatherconditions.

The signal analysis was carried out using a 96-channel (128-channel since July 1997) filter-bankspectrum analyzer with a resolution of 7.5 kHz(0.101 km/s in radial velocity in the 1.35-cm line).A 2048-channel autocorrelator with a resolution of6.1 kHz (0.0822 km/s at 22 GHz) was used startingat the end of 2005. For a pointlike source, an antennatemperature of 1 K corresponds to a flux density of25 Jy.

OH-line observations of W44C at 18 cm wereconducted on the large radio telescope of the Nancay

Radio Astronomy Station of the Paris–Meudon Ob-servatory (France) at various epochs. The telescopeis a Kraus system two-mirror instrument that is ableto observe radio sources near the meridian. Usinga spherical mirror enables tracking of a radio sourceby moving the feed within ±30m/ cos δ in hour anglerelative to the meridian. At declination δ = 0◦, thetelescope beam at 18 cm is 3.5′ × 19′ in right ascen-sion and declination, respectively. The telescope sen-sitivity at λ = 18 cm and δ = 0◦ is 1.4 K/Jy. The noisetemperature of the helium-cooled front-end amplifiersis 35–60 K, depending on the observing conditions.

The spectral analysis was carried out using an8192-channel autocorrelator spectrum analyzer.These channels can be divided into several batteries,each realizing an independent signal analysis in oneof the two main OH lines (1665 and 1667 MHz) inone of four polarization modes. In our observations in2008–2009, the spectrum analyzer was divided intoeight batteries of 1024 channels each. The frequencybandwidth for each battery was 781.25 kHz, andthe frequency resolution 763 Hz. In the 1665 and1667 MHz lines, this corresponds to a radial-velocityresolution of 0.137 km/s. In the 2010 observations,the resolution was twice as good, 0.068 km/s.

A recent upgrade extended the capabilities for po-larization studies [16]. Our 2008–2010 observa-tions appreciably supplement previous polarizationmeasurements obtained using the Nancay radio tele-scope. The radio telescope simultaneously receivestwo perpendicular linear polarizations, which directly

ASTRONOMY REPORTS Vol. 56 No. 1 2012

LONG-TERM MONITORING OF W44C 47

500 Jy

November 30, 1979

December 7, 1979

March 7, 1980

April 25, 1980

706560555045

500 Jy

October 23, 1981

December 16, 1981

February 2, 1982

June 8, 1982

706560555045

June 19, 1980

October 3, 1980

December 16, 1980

February 11, 1981

October 12, 1982

December 29, 1982

March 15, 1983

May 4, 1983

June 8, 1983

February 17, 1981

May 5, 1981

June 9, 1981

October 4, 1983

December 16, 1983

Radial velocity, km/s

Flu

x d

ensi

ty,

Jy

Fig. 2. Catalog of spectra of the H2O maser emission toward G34.3+0.15. The double arrow shows the value correspondingto a division of the vertical axis. The radial velocity is given with respect to the LSR.

yield the intensities of the corresponding linear modes(L0◦, L90◦). Mixing of the signals from the per-pendicular feeds with a phase delay of one quar-ter of a wavelength yields two orthogonal circular-polarization modes (LC, RC). Thus, the three Stokesparameters I, V and Q are actually observed simul-taneously (with an appropriate choice of coordinatesystem). The Stokes parameter U can be measuredafter rotation of the linear-polarization feeds by 45◦.

Combining the polarization modes, we can deriveall four Stokes parameters of the OH maser emis-

sion. The Stokes parameters are determined via theflux densities F in the different polarizations in eachfrequency channel of the spectrum analyzer as follows[16]:

I = F (0◦) + F (90◦) = F (RC) + F (LC),

Q = F (0◦) − F (90◦),

U = F (45◦) − F (−45◦),

V = F (RC) − F (LC).


48 LEKHT et al.

706560555045 65555045Radial velocity, km/s

February 9, 1984

April 4, 1984

June 12, 1984

September 25, 1984

November 13, 1984

December 11, 1984

June 28, 1985

October 17, 1985

November 12, 1985

December 11, 1985

January 3, 1986500 Jy

500 Jy

Flu

x d

ensi

ty,

Jy

December 20, 1984

January 11, 1985

February 5, 1985

March 26, 1985

January 27, 1986

March 11, 1986

June 4, 1986

July 1, 1986

April 17, 1985September 17, 1986

April 11, 1986

60

Fig. 2. (Contd.)

The degree of linear polarization is defined as

mL =

√Q2 + U2

I,

the position angle of the linear polarization is

χ =180◦

πarctan,

(U

Q

)

and the degree of circular polarization is

mC =V

I.

The observational data were processed usingthe GILDAS software package (IRAM, Grenoble,France), which is available at the addresshttp://www.iram.fr/IRAMFR/GILDAS/.



500 Jy

October 31, 1986

November 13, 1986

November 28, 1986

December 26, 1986

706560555045

500 JyDecember 2, 1987

January 6, 1988

February 16, 1988

706560555045

January 29, 1987

February 24, 1987

March 25, 1987

April 26, 1987

March 29, 1988

May 4, 1988

September 20, 1988

October 5, 1988

November 11, 1988

May 28, 1987

November 3, 1987

January 9, 1989

February 2, 1989

Radial velocity, km/s

Flu

x d

ensi

ty,

Jy

Fig. 2. (Contd.)

In the processing, we took into account the effectsof “spurious” polarization due to signal leakage fromone linear-polarization channel to the other. Thetechnique used is described in [17].

Figure 2 presents a catalog of H2O spectra for theperiod from November 1979 to March 2011. Obser-vations were not conducted for technical reasons fromMarch 2006 to December 2007. The double arrowshows the value of a division of the vertical axis inJansky. The horizontal axis plots the velocity withrespect to the Local Standard of Rest (LSR). For

convenience, zero baselines are drawn in the spectra.All the spectra are shown on the same radial-velocityscale.

The catalog of spectra is also shown as a three-dimensional (3D) graph plotting velocity, time, andflux density (Fig. 3). Constructing a 3D image re-quires a uniform grid in the velocity–time plane. Ve-locity intervals are identical, but the time intervals arenot, since the observations were not equally spacedin time. A more uniform grid was obtained by in-troducing additional spectra via a linear interpolation.


50 LEKHT et al.


March 24, 1989

November 27, 1989

December 25, 1989

March 1, 1990

March 27, 1990

April 25, 1990

March 28, 1991

May 11, 1991

March 13, 1992

May 29, 1992

500 Jy 500 Jy

Flu

x d

ensi

ty,

Jy

May 30, 1990

July 4, 1990

November 1, 1990

November 28, 1990

July 7, 1992

October 6, 1992

November 26, 1992

December 16, 1992

December 26, 1990January 19, 1993

January 17, 1992

November 27, 1991

October 10, 1991

June 27, 1991

May 30, 1991

October 23, 1989

September 21, 1989

June 16, 1989

May 30, 1989

April 25, 1989

February 5, 1992

April 15, 1992

Fig. 2. (Contd.)

The bottom graph presents the 3D map for 1980–2005 with a resolution of 0.101 km/s, and the topgraph the 3D map for 2008–2010 with a resolutionof 0.0822 km/s.

Figure 4 presents superpositions of the H2O spec-tra for various time intervals and for the total ob-serving time. The separation was determined by theevolution of the spectra. The intervals are indicatedon each plot.

H2O maser emission is observed within sev-

eral distinct spectral intervals: below 51.6 km/s,51.6–55.0, 55.0–57.9, 57.9–61.7 km/s, and above61.7 km/s. The boundaries between them are deter-mined by a pronounced minima of emission at thesevelocities, shown by the vertical arrows on the lastplot. The most stable emission was observed at 59–61.4 km/s.

The results of the OH-maser observations in the1665 and 1667 MHz lines at various epochs areshown in Fig. 5. The observations were carried out




Flu

x d

ensi

ty,

Jy

March 2, 1993

500 Jy

April 22, 1993

June 24, 1993

September 21, 1993

November 1, 1993

March 20, 1996

October 8, 1996

December 9, 1996

January 21, 1997

February 25, 1997

March 31, 1997500 Jy

February 15, 1994

March 29, 1994

May 17, 1994

August 9, 1994

October 5, 1994

November 9, 1994

April 1, 1997

May 7, 1997

June 6, 1997

July 1, 1997

August 18, 1997July 4, 1995

September 20, 1995

January 24, 1996 September 15, 1997

Fig. 2. (Contd.)

with a resolution of 1 km/s in 1975, 0.137 km/sin 2008, and 0.068 km/s in 2010. The solid anddashed lines show the emission in the left-circularand right-circular polarization, respectively. The ob-serving technique is described in [18].

The OH emission was essentially constant from2008 to 2010. Therefore, we present the Stokes pa-rameters I, Q, U , and V only for the observations withhigher spectral resolution carried out on February 25,2010 (Fig. 6).

3. DISCUSSIONOur analysis of the evolution of the H2O maser

emission during the 30 years of our monitoring hasalso considered spectra obtained on other radio tele-scopes and the results of observations with high an-gular resolution.

3.1. Identification of the Most Stable H2O MaserEmission Features

Using the VLA observations of Fey et al. [4]and Hofman and Churchwell [5], we have identified


52 LEKHT et al.


Flu

x d

ensi

ty,

Jy November 5, 1997

December 24, 1997

January 21, 1998

April 29, 1998

June 3, 1998

July 2, 1998

August 13, 1998

September 29, 1998

October 6, 1998

November 10, 1998

December 16, 1998

1000 Jy

1000 Jy

1780 Ян

January 20, 1999

March 10, 1999

May 12, 1999

August 17, 1999

November 1, 1999

December 16, 1999

February 23, 2000

March 31, 2000

April 25, 2000

Fig. 2. (Contd.)

the main emission features in our monitoring spectrafor epochs close to the VLA observations. No in-terferometric observations of the H2O maser sourcein G34.3+0.15 were carried out during the period ofhigh maser activity, complicating the identification ofsome strong flares. Furthermore, due to the low spec-tral resolution of the VLA observations of Fey et al.[4] (1.3 km/s) , not all emission features are presentin their maps. For instance, the feature responsiblefor the emission at 61.3 km/s, which was observed byus throughout almost all of our monitoring, is absentfrom the VLA images. If this emission belonged to

a cluster other than feature 1, it would certainly havebeen detected by Fey et al. [4].

The most stable H2O emission in W44C, ob-served at radial velocities of 59.6–61.5 km/s, is iden-tified with the main group of maser spots in W44C—cluster 1 [4, 5] (Fig. 1). Figure 7 shows the evolutionof the main emission features of this cluster at radialvelocities of 57.2–61.2 km/s. The filled circles showthe features whose fluxes were highest at the corre-sponding observing epochs, while open circles showthe remaining epochs. In 1998–2003, feature 1 wasappreciably weaker than feature 2. Fluxes are givenalong the curves for the evolution of features 1 and 2.




1290 Jy

Flu

x d

ensi

ty,

Jy

June 14, 2000

July 26, 2000

August 28, 2000

January 17, 2001

February 27, 2001

March 27, 2001

April 11, 2001

May 29, 2001

September 4, 2001

October 25, 2001

1000 Jy1000 Jy

November 27, 2001

December 18, 2001

February 5, 2002

March 26, 2002

April 23, 2002

July 23, 2002

August 13, 2002

September 16, 2002

October 8, 2002

November 11, 2002

December 19, 2002

Fig. 2. (Contd.)

The triangles mark the positions of emission featuresfrom the VLA observations of Fey et al. [4]. The hori-zontal line segments show time intervals when strongflares occurred at radial velocities of 53–56 km/s; theparameters of these flares (radial velocities and fluxes)are given.

Features 1–3 are arranged radially with respectto the central star in W44C, in order of increasingradial velocity [4]. Some other emission features areidentified with clusters of maser spots 2, 3, and 9(Fig. 1), which are also located near the shock front.

3.2. Long-Term Evolution of the H2O Maser

The position of the peak of the main emission ofcluster 1 varied from 59.5 to 60.5 km/s in a com-plicated manner. Valdettaro et al. [10] showed thatthis emission is fairly stable. Our 30-year monitoringlikewise indicates this emission to be stable, in thesense that it was observed throughout our moni-toring. However, its evolution was complex. Wehave selected several components. The main ones(1 and 2) were observed throughout our monitoring,with the exception of 1979–1981 for feature 2. Theintensity of each component separately varied fairly


54 LEKHT et al.


Flu

x d

ensi

ty,

Jy January 28, 2003

March 25, 2003

April 23, 2003

June 26, 2003

August 5, 2003

September 23, 2003

December 17, 2003

January 28, 2004

March 10, 2004

April 21, 2004

1000 Jy

1000 Jy

May 25, 2004

June 16, 2004

July 22, 2004

November 4, 2004

December 24, 2004

April 14, 2005

June 28, 2005

September 27, 2005

March 28, 2006

Fig. 2. (Contd.)

strongly: by about a factor of five for feature 1 and bymore than a factor of 20 for feature 2. The variations ofthe integrated fluxes of these features were consider-ably smaller. Thus, we may consider the H2O maseremission source W44C to be fairly stable.

The highest activity of the H2O maser took placein 1987–1988 and in 2001. This activity was mini-mum in 1980, 1992, and 2009, and the minima weremore spread out in time than the maxima. Never-theless, we suggest that the maxima alternated withan interval of 14 years and minima with intervals of11.5 and 17.5 years. A formal calculation of the meanperiod of the integrated-flux variability yields a valueof ∼14 years.

3.3. H2O Flare EmissionFlares of individual emission features occurred

fairly frequently. However, strong flares (series offlares) exceeding 800 Jy were observed in only threetime intervals (Fig. 4), at velocities of 52–57 km/s.

In the first interval (1986–1988), the flux densityin flares reached 1800 Jy. At that time, the structureof the entire line profile changed, while the emissionof the main components, 1 and 2, was preserved. Itis important that, after the first series of flares, theradial-velocity drift of component 2 was small. Thedrift of component 1 is approximated well by a sinewave within 0.7 km/s. The main components of theflare are identified with the main cluster 1. All this



6050403020Radial velocity, km/s

February 6, 2008

400 JyMay 13, 2008

April 15, 2009

June 16, 2009

August 24, 2009

November 9, 2009

December 14, 2009

January 26, 2010

February 25, 2010

March 30, 2010

April 27, 2010

August 23, 2010

October 11, 2010

November 1, 2010

December 15, 2010

February 7, 2011

March 3, 2011

March 30, 2011

Flu

x d

ensi

ty,

Jy

Fig. 2. (Contd.)

could indicate a global character of the activity of theH2O maser in W44C.

The flares in 2000–2002 were less prolonged. Thisemission is most likely identified with another cluster(cluster 2). The flux density in one of the flaresreached 1900 Jy. At that time. no structural changesof the main emission of cluster 1 (velocity or flux den-

sity) took place. The flares probably had a local char-acter, but occurred during a period of high activity of

the entire maser source. The radial-velocity and fluxvariations of the two main components of this flare are

shown in Fig. 8. The radial-velocity drift in the local

rest frame is complex, and is approximated well by the


56 LEKHT et al.

8060

4020

2011201020092008

800

600

400

200

0

Flu

x d

ensi

ty, Jy

Years Radial

velocit

y, km/s

2400

2100

1800

1500

1200

900

600

300

0

Flu

x d

ensi

ty, Jy

1980 19851990

19952000 2005

Years

70

65

60

55

50

45

Radial velocity

, km/s

Fig. 3. Three-dimensional representation of the catalog of H2O maser spectra in G34.3+0.15.

fitted curves. An approach and subsequent recessionof the features in the spectrum can be clearly traced.

Some weaker and shorter flares occurred duringthe monitoring, in a fairly broad interval of radialvelocities.

The observed variability of the components couldbe a consequence of turbulent motions of matter onscales comparable to the sizes of maser spots andclusters of maser spots.

3.4. OH Emission

Owing to its one unpaired electron, the OHmolecule has a large magnetic dipole moment,1.66 Debye [19]. Therefore, the observed OH linespectrum is strongly influenced by an external mag-netic field. In a magnetic field, the main lines at1665 and 1667 MHz split into three components: anunshifted π component with linear polarization andtwo σ components displaced upward and downwardin frequency that are elliptically polarized in opposite



70656055504540

1200

800

400

0

0

400

0

400

800

0

400

800

1200

1600

0

400

800

0

400

800

1200

800

I.1993−X.1998

1990−1992

V.1998−XII.1989

I.1986−III.1988

1984−1985

1979−1983

(a)

(b)

(c)

(d)

(e)

(f)

2000

Flux density, Jy

0

1600

1200

800

400


1979−2011 (j)

70656055504540

2000

1600

1200

800

400

01600

1200

800

400

0800

400

0

XI.1998−IX.2002

X.2002−III.2006

2008−2011

(g)

(h)

(i)

Fig. 4. Superposition of H2O emission spectra for various time intervals (indicated on the graphs) and for the entire monitoringtime (lower right-hand graph). All spectra are given on the same scale.

directions. In a longitudinal field with intensity B, thefrequency separation between the σ components is

Δν =52

gJμ0

hB (1)

for the 1665 MHz line and

Δν =32

gJμ0

hB (2)

for the 1667 MHz line, where gJ is the Lande factor(0.935 for both lines), μ0 is the Bohr magneton, his Planck’s constant, μ0/h = 1.39967 kHz/mG, andthe separation of the σ components in radial veloc-ity is 0.590 km/(s mG) for the 1665 MHz line and0.354 km/(s mG) for the 1667 MHz line [20].

Measuring the velocity difference of the oppositelypolarized σ components, we can determine the inten-sity of the longitudinal component of the magnetic


58 LEKHT et al.

686460565248

20

15

10

5

0

1665 MHz

January 17 and 29, 1975

RL

686460565248

30

20

10

0

1667 MHz

R

L

6260585654

200

150

100

50

0200

150

100

50

0

150

100

50

0

150

100

50

0

Flu

x d

ensi

ty,

Jy

January 6, 2008

February 25, 2010

April 6, 2010

July 4, 2010

300

200

100

0300

200

100

0300

200

100

0300

200

100

0


Fig. 5. Spectra of OH maser emission in the 1665 and 1667 MHz lines for left-circular (L) and right-circular (R) polarizationsfor various observing epochs. The spectral resolution was 1 km/s on January 17 and 29, 1975; 0.138 km/s on January 6, 2008;and 0.068 km/s in 2010.

field. The positions of the masers of a Zeeman pair onmaps measured in the different polarizations coincideto within the errors (for the VLBI measurements,to within fractions of a milliarcsecond). This con-dition follows from the nature of Zeeman splitting:

both σ components are radiated by each moleculesimultaneously as a result of its precession in themagnetic field. VLBI observations of masers confirmthat there is indeed positional coincidence of the σcomponents, though their intensities are usually not



200

150

100

50

0

60

40

20

0

0

−20

−40

100

50

0

−50646260585654

1665 MHzI

Q

U

V

Flu

x d

ensi

ty,

Jy


50

0

−50

−100

−150

−200

−200

−100

0

0

50

100

0

100

200

300

400

1667 MHzI

Q

U

V

Fig. 6. Stokes parameters for the 1665 and 1667 MHz OH emission on February 25, 2010.

equal. Basically, the total or partial suppression of oneσ component by the other is possible. In this case,only one σ component with a high degree of circularpolarization (up to 100%) is observed.

The general profile of the 1665 and 1667 MHz OHlines toward source C is formed mainly by emissionfrom the regions near the main clusters of H2O spots,namely, 1, 2, and 3. Thus, the most intense H2Oand OH maser emission arises in spatially nearbyregions. As was noted in the Introduction, the mostintense OH sources are located just upstream of theionization front of the HII region W44C [6, 7].

Figures 5 and 6 present our 1665 and 1667 MHzOH line data obtained on the Nancay radio tele-scope. Figure 9 shows the evolution of the 1665 and1667 MHz emission in W44C in 1970–2010 accord-ing to [6–9, 14, 16, 21–23] and our own Nancayobservations in 1974, 1975, and 2008–2010. Sincethe onset of OH line observations of W44C [21], thepeak flux density of the most intense feature in the1667 MHz line, at 58 km/s, has increased by morethan an order of magnitude. The flux density in the1665 MHz line has also increased. One peculiarityof W44C that distinguishes it from other OH masers


60 LEKHT et al.

2010200520001995199019851980Years

54.2: 132053.3: 207054.1: 1920 55.7: 1900

61.5

57.0

61.0

60.5

60.0

59.5

59.0

58.5

58.0

57.5

Rad

ial v

elo

city

, km

/s

2

1

3

4

100400

900

1800

750

12501750

640

280

400

1000

400

850

550

800

350

1150

1000900

370

700

400

500

340

200�300

480 300

180

700

5

Fig. 7. Evolution of H2O emission features of the main cluster in the radial-velocity interval 57.2–61.2 km/s. The filled circlesshow features with the greatest flux densities at the corresponding observing epochs, and open circles features for the remainingepochs. Flux densities are given along the graphs of the evolution of features 1 and 2. The triangles show the positions of theemission features observed with the VLA by Fey et al. [4]. The horizontal line segments show the time intervals for strongflares at velocities of 53–56 km/s; the parameters of the flares (radial velocities in km/s and flux densities in Jansky) are given.

in star-forming regions is the greater intensity of the1667 MHz line compared to the 1665 MHz line. Inthe majority of masers of this class, the 1665 MHzline is more intense than the 1667 MHz line. Thisdifference may be related to the particulars of the OHmaser pumping in W44C near the shock front of thecometary HII region.

4. MAIN RESULTS

Let us summarize the main results obtainedfrom our 30-year monitoring of the water maserand long-term monitoring of the hydroxyl maser inG34.3+0.15.

1. We present here a catalog of H2O maser spectrain the 1.35 cm line toward G34.3+0.15 for the period

from November 1979 to March 2011 (Fig. 2), witha mean interval between observational sessions ofabout two months. The radial-velocity resolution was0.101 km/s before and 0.0822 km/s after the end of2005.

2. We have found an alternation of maxima andminima of the H2O maser activity. A formal calcu-lation of the mean period of activity yields ∼14 years;however, this is consistent with our results for a num-ber of other sources associated with regions of activestar formation.

3. We have observed two series of strong flaresof the H2O maser emission, also with an interval of14 years, which were associated with two differentclusters of maser spots (1 and 2) localized at the



200420032002200120001999Years

58

57

56

55

1600

1200

800

400

0

Rad

ial v

elo

city

, km

/sF

lux

den

sity

, Jy

6

7

7

6

Fig. 8. Evolution of the main H2O emission features of cluster 2 during the flare of 2000–2002. The radial-velocity variationsare approximated by curves.

shock front at the periphery of the ultracompact re-gion UH II. These series of flares took place duringperiods of enhanced maser activity, and are probablyrelated to cyclic activity of the protostellar object inUH II (component C). In the remaining time inter-vals, there were mainly short-lived flares of singlefeatures.

4. The observed character of the variability of theemission of individual H2O maser components maybe a consequence of turbulent (including vortical)motions of matter, within maser spots and clusters ofspots located near the shock front.

5. Since the discovery of OH emission in W44Cat the beginning of the 1970s, the flux density of theOH lines has gradually increased, and has currentlyreached a maximum for the entire time covered by theobservations. This increase could be related to thepropagation of a shock exciting maser emission deepwithin the molecular cloud.

6. We have estimated the intensity of the line-of-sight magnetic field to be −1.2 mG from the splittingof the 1667 MHz OH maser feature at 58.6 km/s. Thefield is directed toward the observer, consistent with


62 LEKHT et al.

201020052000199519901985198019751970Years

400

0

300

200

100

Flu

x d

ensi

ty,

Jy1667 MHz

1665 MHz

Fig. 9. Evolution of the 1665 and 1667 MHz OH emission in W44C in 1970–2010 (data of [6–9, 14, 16, 21–23] and ourobservations on the Nancay radio telescope in 1975 and 2008–2010, open symbols).

the general pattern of the magnetic field in the W44Cregion mapped by interferometric observations [9].

ACKNOWLEDGMENTS

This work was supported by the Russian Founda-tion for Basic Research (project code 09-02-00963-a). The authors are grateful to the staff of thePushchino (Russia) and Meudon (France) observa-tories for their great help with the observations.

REFERENCES1. M. J. Reid and P. T. P. Ho, Astrophys. J. 288, 417

(1985).2. D. O. S. Wood and E. Churchwell, Astrophys.

J. Suppl. Ser. 69, 831 (1989).3. B. E. Turner and R. H. Rubin, Astrophys. J. 170,

L113 (1971).4. A. L. Fey, R. A. Gaume, G. E. Nedoluha, and

M. J. Claussen, Astrophys. J. 435, 738 (1994).5. P. Hofner and E. Churchwell, Astron. Astrophys.

Suppl. Ser. 120, 283 (1996).6. A. L. Argon, M. J. Reid, and K. M. Menten, Astro-

phys. J. Suppl. Ser. 129, 159 (2000).7. X. Zheng, J. M. Moran, and M. J. Reid, Mon. Not.

R. Astron. Soc. 317, 192 (2000).8. X. Zheng, M. J. Reid, and J. M. Moran, Astron.

Astrophys. 357, L37 (2000).9. N. Gasiprong, R. J. Cohen, and B. Hutawarakorn,

Mon. Not. R. Astron. Soc. 336, 47 (2002).10. R. Valdettaro, F. Palla, J. Brand, et al., Astron. Astro-

phys. 383, 244 (2002).

11. R. Cesaroni, C. M. Walmsley, C. Kompe, andE. Churchwell, Astron. Astrophys. 252, 278 (1991).

12. E. Churchwell, C. M. Walmsley, and D. O. S. Wood,Astron. Astrophys. 253, 541 (1992).

13. R. A. Gaume, A. L. Fey, and M. J. Claussen, Astro-phys. J. 432, 648 (1994).

14. M. I. Pashchenko, Sov. Astron. Lett. 1, 241 (1975).15. E. E. Lekht, M. I. Pashchenko, G. M. Rudnitskii, and

R. L. Sorochenko, Sov. Astron. 26, 168 (1982).16. M. Szymczak and E. Gerard, Astron. Astrophys. 423,

209 (2004).17. V. I. Slysh, M. I. Pashchenko, G. M. Rudnitskii, et al.,

Astron. Rep. 54, 599 (2010).18. M. I. Pashchenko, G. M. Rudnitskii, and P. Colom,

Astron. Rep. 53, 541 (2009).19. A. A. Radzig and B. M. Smirnov, Reference Book

on Atomic and Molecular Physics (Atomizdat,Moscow, 1980) [in Russian].

20. R. D. Davies, in Galactic Radio Astronomy, Pro-ceedings of the IAU Symposium No. 60, Maroochy-dore, Queensland, Australia, Sept. 3–7, 1973, Ed.by F. J. Kerr and S. C. Simonson (Reidel, Dordrecht,Holland, Boston, 1974), p. 275.

21. B. E. Turner, Astron. Astrophys. Suppl. Ser. 37, 1(1979).

22. J. L. Caswell and R. F. Haynes, Austral. J. Phys. 36,417 (1983).

23. R. A. Gaume and R. L. Mutel, Astrophys. J. Suppl.Ser. 65, 193 (1987).

Translated by G. Rudnitskii


long-term monitoring of the w44c (g 34.3+0.15) maser in the 1.35 cm water vapor and 18 cm hydroxyl...

Documents