Free Access
Issue
A&A
Volume 529, May 2011
Article Number A95
Number of page(s) 12
Section Interstellar and circumstellar matter
DOI https://doi.org/10.1051/0004-6361/201116648
Published online 11 April 2011

© ESO, 2011

1. Introduction

Maser observations have been able to provide important information on the magnetic field strength and structure in the densest areas of massive star forming regions and other astrophysical objects (for a review see e.g. Vlemmings 2007). As masers probe different densities and conditions, such as outflows, disks and shocks, combining the various maser observations can provide a detailed picture of the magnetic field that is needed for a proper understanding of these complex regions. The majority of the magnetic field information still comes from OH masers. These typically display a few mG field strength (e.g. Fish & Reid 2006; Bartkiewicz et al. 2005) and a coherent magnetic field structure. For example the polarization of the OH masers of W75N matches a toroidal field in a massive torus or disk (e.g. Hutawarakorn et al. 2002; Gray et al. 2003). The H2O masers mostly probe shock compressed fields and occasionally imply enhanced magnetic fields due to a nearby protostellar dynamo (e.g. Vlemmings et al. 2006a; Surcis et al. 2011). However, because of the shocked nature of the H2O masers and the potential effect of Faraday rotation on the linear polarization of the OH masers, they are not necessarily the best probes of the magnetic field structure.

Recently, it has been discovered that methanol masers, one of the most numerous massive star formation maser species, also displays significant linear and circular polarization (e.g. Ellingsen 2002; Vlemmings et al. 2006b; Dodson 2008; Vlemmings 2008; Sarma & Momjian 2009). As methanol masers are less affected by Faraday rotation because of their higher frequency and as the different methanol maser transitions probe different areas of massive star formation, their potential in magnetic field studies is enormous. Unfortunately, the interpretation of maser polarization depends critically on the Zeeman frequency shift in relation to the maser saturation level (e.g. Watson 2009) and determination of these quantities heavily relies on several assumptions. In this paper we present new observations of the splitting between the right- and left-polarization spectra of 6.7 GHz methanol masers to expand the sample presented in Vlemmings (2008, hereafter Paper I). We observed most of the strong northern- and southern-hemisphere 6.7 GHz masers, although splitting could only be detected in the northern sample because of calibration issues. While the splitting is attributed to the Zeeman effect, we show that an error was introduced in earlier Zeeman splitting calculations that makes the determination of magnetic field strength from the measured splitting highly uncertain.

2. Observations and analysis

2.1. Effelsberg observations

The observations of the northern maser sample were taken on June 17−19th and June 27th 2008 using the 5 cm primary focus receiver of the 100-m Effelsberg1 telescope. The full width at half-maximum (FWHM) beam of the telescope is  ~2 arcmin at the maser frequency. The setup was identical to that used in Paper I. We used a position-switch mode with a 2 min cycle time and the data were collected using the fast Fourier transform spectrometer (FFTS) with two spectral windows, corresponding to the right- and left-circular polarizations (RCP and LCP). The spectral windows of 20 MHz were divided in 16   384 spectral channels, resulting in a  ~0.055 km s-1 channel spacing and were centered on the local standard of rest (LSR) source velocities.

The data were reduced using the Continuum and Line Analysis Single-dish Software (CLASS) package and the amplitudes were calibrated using scans on 3C 286. As some of the data was taken with the same setup and at the same date of the monitoring observations of G9.62+0.20 (Vlemmings et al. 2009, hereafter V09), we were able to compare the fluxes with simultaneous Hartebeesthoek radio telescope observations of that source (Goedhart, priv. comm.). From this we estimate the absolute flux errors to be  ~10%. Any larger flux variations are likely due to maser variability and/or intrinsic changes of the maser structure.

2.2. Parkes observations

The southern maser sample was observed with the Parkes 64-m telescope using the methanol multibeam (MMB) receiver between Aug. 15 − 18th 2008. The MMB system is detailed in Green et al. (2009). Only two beams of the receiver were used, performing beam switching with a 2 min cycle time. We observed two windows, centered on the LSR source velocity for the 6.7 GHz methanol maser transition and for the 6.035 GHz excited OH maser transition respectively. The OH maser observations will be presented in a forthcoming paper (Torres et al., in prep.). The FWHM of the Parkes beam at 6.7 GHz is 3.2 arcmin. Using the spectrometer with 2048 channels and a bandwidth of 4 MHz, we reach a spectral resolution of  ~0.09 km s-1.

The data calibration was performed using the single-dish ATNF Spectral Analysis Package (ASAP) embedded in the Common Astronomy Software Applications (CASA) package. Unfortunately, the scan-to-scan stability of the RCP- and LCP-signals turned out to be insufficient to determine the Zeeman splitting. Although the exact cause of this is unclear, the most likely explanation is a lack of baseline stability on short time scales and potential narrow baseline ripples that cannot be removed by a low polynomial baseline fit. Thus, the further analysis in the paper is focused on the Effelsberg observations. For completeness, the total intensity maser spectra and derived peak and integrated fluxes are presented in the online material, Fig. 7, and Table 2.

2.3. Source selection

The Effelsberg sample was taken from the 6.7 GHz methanol maser catalogue by Pestalozzi et al. (2005) and consists of the strongest (>50 Jy) northern maser sources observable from Effelsberg. The first part of the sample was presented in Paper I but is also presented here for completeness. As a number of the masers display significant difference in the catalogue flux and the observed flux, several of the sources presented here fall below the 50 Jy limit imposed during source selection. The Parkes sample was taken from the same catalogue but limited to the Southern sources with a listed flux  >100 Jy. To detect the Zeeman splitting the goal was to reach a signal-to-noise of  >3000 and thus the total integration time per source was variable. When an initial scan indicated the observed flux was significantly lower then the catalog flux and as a result a Zeeman splitting detection would be unlikely within reasonable time, integration time was typically shortened. Thus, the noise level varies from source to source and ranges from 20 mJy to 100 mJy for each polarization.

2.4. Analysis method

As in Paper I and V09, we used the cross-correlation between the RCP and LCP signals to determine the Zeeman splitting of the methanol masers. The advantage of this method is, that it is insensitive to relative gain calibration errors between the two polarization. In the case of methanol Zeeman splitting this is specifically important, as the splitting is typically of order 0.5 m s-1. The circular polarization arising from such small splitting is  < 0.5% (Paper I).

2.5. Error analysis

The errors on the Zeeman splitting determination depend on the rms noise in the RCP and LCP maser spectra. However, it was shown in V09, that for masers stronger than 50 Jy beam-1, the noise in the channels with maser emission increases. We have thus corrected the Zeeman splitting values determined in Paper I following the relation found in V09 between rms channel noise and maser flux. Note that the error estimates are thus very conservative, as we have taken the rms noise in the channel with the strongest maser emission to determine the total Zeeman splitting error.

3. Methanol maser Zeeman splitting

3.1. Uncertainty in the splitting coefficient

The methanol molecule is a non-paramagnetic molecule and as a result the Zeeman splitting under the influence of a magnetic field is extremely small. The split energy, ΔEZ, of an energy level under the influence of a magnetic field, B, can be described as ΔEz = gLμNMJB, where MJ denotes the magnetic quantum number for the rotational transition described with the total angular momentum quantum number J, B is the magnetic field strength in units of Tesla ( = 104 G), μN is the nuclear magneton and gL is the Landé g-factor. The Zeeman effect is determined by the Landé g-factor, which needs to be determined from laboratory spectroscopy.

In the previous publications of methanol polarization, the g-factor used to determine the magnetic field strength was based on laboratory measurements performed many years ago on a number of methanol transitions near 25 GHz methanol (Jen 1951). He found empirically that the methanol g-factor, could be described by the equation: (1)However, there are several caveats regarding these measurements. Firstly, gL is an average of the true g-factor of several interacting states. Additionally, the measurements are classified as preliminary in Jen (1951), and the exact transitions that were used are not specified. The observations were done on poorly identified transitions around 25 GHz with ΔJ = 0 and K = 2 − 1, which likely indicates it concerns the E1-type methanol maser. It is thus not impossible that an extrapolation to the 6.7 GHz 51 − 60 A+ methanol transition and others transitions with different ΔJ and quantum number K is invalid.

While it is thus unclear if the g-factor determined in 1951 can be used for the 6.7 GHz methanol maser, it is the only estimate available to us at the moment. Using this, Vlemmings et al. (2006b) determined the Zeeman splitting coefficient to be 0.049 km s-1 G-1. After a reanalysis, it was recently found that an unfortunate calculation error was introduced in that paper. The actual Zeeman splitting coefficient for the 51 − 60 A+ 6.7 GHz methanol maser transition extrapolated from the laboratory measurements is 0.005 km s-1 G-1, an order of magnitude smaller than previously determined. This error has also affected the magnetic field calculations of the other methanol maser transitions (e.g. Sarma & Momjian 2009).

If we can still attribute the measured splitting between the RCP- and LCP-spectra of the 6.7 GHz methanol maser as standard Zeeman splitting, the implication of the new Zeeman splitting coefficient is that the magnetic field strength in the methanol maser region is an order of magnitude larger. Based on the results in Paper I, this would thus imply that the magnetic field is of order 100 mG. In that case, the magnetic field measured on the methanol masers is almost two orders of magnitude stronger than that measured on OH masers (e.g. Fish & Reid 2006). Considering the empirically determined scaling relation (Crutcher 1999) and the indication that the methanol masers exist in gas with a density at most an order of magnitude larger that of the OH masers (e.g. Green et al. 2007), such a large difference between the methanol and OH determined magnetic field strength is surprising. This seems to imply that the extrapolated g-factor is uncertain by an order of magnitude or potentially that instrumental or other non-Zeeman effects dominate the 6.7 GHz splitting measurements.

3.2. Potential non-Zeeman effects

As the best estimate for the 6.7 GHz Zeeman coefficient is much smaller than previously assumed, the determined magnetic fields based on the regular Zeeman effect are much larger than would be expected. We thus need to again determine if non-Zeeman effects could be the cause of the observed splitting between the RCP- and LCP-signal.

First we need to consider instrumental effects. The 6.7 GHz methanol maser splitting was first measured with the 100-m Effelsberg telescope (Paper I). In our monitoring observations (V09), we have shown the observed splitting to be reproducable and the observing system to be robust. The splitting of the masers in W75N was later confirmed in high-angular resolution European VLBI Network (EVN) observations (Surcis et al. 2009). Additionally, recent observations with the Hobart 26-m antenna confirmed the splitting of G09.62+0.20 (Ellingsen, priv. comm.). Therefore instrumental effects can likely be ruled out.

Another effect is the propagation of strong linear polarization that can cause circular polarization when the direction of the magnetic field changes significantly along the maser propagation direction (Wiebe & Watson 1998). For a smooth change of magnetic field direction of  ~1 rad along the maser, the fractional circular polarization caused by this effect is approximately , where ml is the fractional linear polarization. The typical polarization observed in high angular resolution observations is 1 − 4% (e.g. Vlemmings et al. 2006b; Dodson 2008; Surcis et al. 2009), indicating that this effect contributed at most  ~0.04%, a fraction of the observed values.

Finally, a potentially important effect is caused by a rotation of the axis of symmetry for the molecular quantum states. This can occur when, as the maser brightness increases while it becomes more saturated, the rate for maser stimulated emission R becomes larger than the Zeeman frequency shift gΩ. While gΩ ≫ R, the magnetic field direction is the quantization axis. Then, when R becomes larger than gΩ, the molecules interact more strongly with the radiation field than with the magnetic field and the quantization axis changes towards the maser propagation direction. This change will cause an intensity-dependent circular polarization that mimics the regular Zeeman splitting. This has been shown by Nedoluha & Watson (1990a) for a J = 2 − 1 transition although the effect decreases for transitions with higher angular momentum such as the 6.7 GHz methanol maser and the exact magnitude is hard to determine.

From the recalculated Zeeman splitting coefficient derived above, gΩ ≈ 0.1B [mG ]  s-1 for the 6.7 GHz methanol maser. The rate for stimulated emission can be estimated using: (2)Here A is the Einstein coefficient for the maser transition, which is equal to 0.1532 × 10-8 s-1 (Cragg et al. 1993), and k and h are the Boltzmann and Planck constants respectively. The maser frequency is denoted by ν, and Tb and ΔΩ are the maser brightness temperature and beaming solid angle. The ratio between zeeman splitting and rate of stimulated emission for the 6.7 GHz methanol masers, assuming the recalculated g-factor, can thus be given by: (3)Our observations indicate typically Tb ≲ 1010 K (Surcis et al. 2009; Vlemmings et al. 2010), although for the brightest masers Tb ~ 1012 (Minier et al. 2002). The beaming angle ΔΩ is harder to estimate and decreases rapidly with increasing maser saturation level. For H2O masers, beaming angles have been estimated to be of order 10-5 − 10-4 (Nedoluha & Watson 1991). If we very conservatively assume a maser beaming angle of ΔΩ ≈ 10-2, the typical maser stimulated emission R ~ 0.04 s-1, and for the most saturated masers R ≲ 4 s-1, even without considering the rapid decrease of ΔΩ due to the fact that beaming becomes much more pronounced for the stronger masers. Thus, typically, gΩ/R > 1 and only for the most saturated masers would we expect the non-Zeeman effect to be applicable. As detailed in V09, this could be the cause for the circular polarization variability seen during the flare of G09.62+0.20. Another indication that non-Zeeman effects are not the cause of the observed splitting is the lack of an observed relation between maser flux and the RCP- and LCP-splitting illustrated in Fig. 1. Even ignoring the strongest masers, W3OH and G09.62+0.20, due to their complicated maser structure and flaring nature respectively, it is clear that the observed splitting is independent of maser flux.

thumbnail Fig. 1

Observed splitting between the right- and left-polarization spectra vs. the maser peak flux. The lack of an increase of the splitting with flux argues against a non-Zeeman interpretation of the observed splitting.

Open with DEXTER

Thus, there are no indications of a non-Zeeman contribution to the observed splitting, which we consequently still attribute to actual Zeeman splitting. However, the exact Zeeman splitting coefficient is highly uncertain and we cannot determine the exact field strength. Based on the measurements of W75N, where, assuming a splitting coefficient of 0.049 km s-1 G-1, the methanol maser magnetic field is identical to the field measured during an OH maser flare (Surcis et al. 2009), we suggest that the originally assumed Zeeman splitting coefficient is closer to the actual value. New laboratory measurements are needed to settle this issue. Thus, the derived magnetic field strengths could be different by an order of magnitude. However, with the non-Zeeman effect unlikely to be in effect, linear polarization measurements still are excellent probes of the magnetic field morphology, especially as the linear polarization fraction appears to be too low to be due to anisotropic pumping even for such small gΩ (Nedoluha & Watson 1990b). Additionally, as gL is a signed quatity, with its sign not in doubt, the Zeeman splitting measurements still reveal the direction of the magnetic field, with negative Zeeman splitting values indicating a field oriented towards the observer.

4. Results

4.1. Zeeman splitting

The results of our survey are presented in Table 1. The table list the source name, possible alternate names, the central VLSR velocity, peak and integrated fluxes and the measured Zeeman splitting ΔVz. This value of the splitting corresponds to the flux averaged Zeeman splitting of the entire maser spectrum and can be used to determine the line-of-sight magnetic field strength (B||) using the Zeeman coefficient of 0.005 m s-1 G-1. As described in Sect. 3 however, this Zeeman coefficient value is highly uncertain and is potentially wrong by an order of magnitude. The table lists the sources from Paper I with the increased error bars. For the sources with significant changes of the Zeeman splitting between different spectral features, we list multiple values. Total intensity spectra and Zeeman splitting determined across the spectrum are presented in online Fig. 5. The total intensity spectra of the sources with no Zeeman splitting detection are shown in online Fig. 6. The sources observed with the Parkes telescope are listed with VLSR velocity and peak and integrated fluxes in online Table 2. The spectra are shown in online Fig. 7.

Table 1

Zeeman splitting results.

thumbnail Fig. 2

Total intensity spectrum (bottom) and Zeeman splitting (top) for G111.43+076 (NGC 7538). The red solid line and square symbols denote the measurements presented in Paper I, taken at Nov. 12th 2007, the black dashed line and solid circles are the measurements presented here.

Open with DEXTER

4.2. Comparison between observing epochs

Although the stability of the Effelsberg polarization measurements was already confirmed by the monitoring observations of G09.62+0.20 presented in V09, we performed additional tests, observing 5 sources that were previously observed in Paper I. This included three sources for which we had only determined an upper limit to the Zeeman splitting. As can be seen in Table 1, the flux averaged Zeeman splitting of both G12.89-0.49 (IRAS 18089-1732) and G111.53+0.76 (NGC 7538) are fully consistent at the two epochs. As shown in Fig. 2, also the variation of Zeeman splitting across the maser spectrum of G111.53+0.76 is reproduced. The only difference is the non-detection of the field around  − 56 km s-1, which is due to the fact that we spent slightly less observing time on this source in the second epoch and thus cannot confirm the earlier detection, that was close to 3σ, in this velocity interval. In the observations presented here, we were able to determine the Zeeman splitting on the other three sources (G25.71+0.04, G32.03+0.06 and G43.80-0.13) by increasing the observing time. These values were in agreement with the previously determined upper limits for G32.03+0.06 and G43.80-0.13, while for G25.71+0.04 the current value should have been detected at the 4σ level. As the other sources indicate good stability, we thus conclude that for this specific source a small intrinsic change has occurred in the flux averaged field strength.

5. Discussion

thumbnail Fig. 3

Distribution of observed Zeeman splitting (thick solid line) in the Effelsberg 6.7 GHz methanol maser sample. Over-plotted are models for the line of sight Zeeman splitting in the case of where the magnetic field is taken to have a Gaussian distribution with a Zeeman splitting dispersion of σZ in each dimension. Following a simple analysis, the best fitted model has σZ = 0.4 m s-1 which would corresponds to a full 3-dimensional equivalent Zeeman splitting of  ⟨ ΔVz ⟩  ≈ 0.62 m s-1. As an indication, using the uncertain laboratory extrapolated g-factor (Sect. 3) this corresponds to  ⟨ |B| ⟩  ≈ 120 mG.

Open with DEXTER

Although we attribute the observed splitting between the RCP- and LCP-spectra as true Zeeman splitting we cannot derive a meaningful magnetic field strength. We plot the distribution of measured Zeeman splitting in Fig. 3. To obtain an indication of the total magnetic field, we assume the field to have a random orientation with each one dimensional field component described by a Gaussian distribution with dispersion σZ. We then perform Monte-Carlo modeling of the observed Zeeman splitting distribution using a description of the detection limit that describes the error bars of our observations. Three of these distributions are also shown in Fig. 3. We do not perform any further statistical tests, as the Zeeman coefficient uncertainty will dominate any final results and the described magnetic field model is unlikely to be realistic, but the best fit to the observed distribution is found for σZ ≈ 0.4 m s-1. If the laboratory value can be extrapolated to the 6.7 GHz methanol masers this would indicate a rather high average field strength of  ⟨ |B| ⟩  ≈ 120 mG. For comparison, the typical field strength measured in OH masers |BOH| ≈ 5 (Fish & Reid 2006). If methanol masers exist in gas with H2 number densities of order 108 cm-3 (Cragg et al. 2005), at most an order of magnitude more than the typical density of OH masers, the magnetic field vs. density relation from Crutcher (1999) (B ∝ n0.47) would predict an average methanol field strength |Bmeth| ≈ 15 mG. This would be consistent with the values derived from the Zeeman splitting only if the Zeeman splitting coefficient for the 6.7 GHz A-type methanol is an order of magnitude larger than the value extrapolated from the 25 GHz E-type methanol Landé-factor measurements.

thumbnail Fig. 4

Magnetic field direction derived from the methanol maser Zeeman splitting observations presented in this paper projected onto the Galactic plane. The symbols are the observed star forming region with kinematic distances from Pestalozzi et al. (2005) unless better distances were available. The open circles and crosses indicate a clockwise and counterclockwisemagnetic field direction respectively. The approximate location of the spiral arms is indicated as taken from Taylor & Cordes (1993).

Open with DEXTER

Despite the uncertainty in the magnitude of the Landé factor, the sign of Zeeman splitting is not in question. The observations thus represent a large set of line-of-sight magnetic field directions toward massive star forming regions. It has been suggested that the magnetic field direction measured in interstellar OH masers might be linked to the overall Galactic magnetic field (Davies 1974). As we have previously shown in Paper I that the sign of the methanol maser derived magnetic field direction corresponds well with that derived from OH masers, as similar link might thus exist between the field direction determined from methanol masers and the Galactic magnetic field. In Fig. 4 we show the magnetic field direction measured for our complete sample projected onto the Galactic plane. The observations show that the vast majority of the masers display a clockwise oriented magnetic field consistent with Galactic rotation. However, one would have to question if the densest regions of star formation retain a magnetic field aligned with the Galactic field through compression and likely rotation. A more thorough analysis of OH maser Zeeman splitting results does not support such a relation (Fish et al. 2003), and also more recent work based on dust linear polarization observations find no clear relation between the large scale Galactic magnetic field structure and that measured in star forming regions (Stephens et al. 2011). Still, the clear preference for a clockwise field orientation in the methanol maser regions is suggestive of a global trend. A more detailed analysis is however beyond the scope of this paper.

6. Concluding remarks

This paper presents Zeeman splitting measurements obtained on a flux limited sample of 6.7 GHz methanol masers. The observations of this unique tracer of massive star formation were performed with the Effelsberg and Parkes radio-telescopes. Unfortunately, the stability of the Parkes observations precluded the determination of Zeeman splitting and our sample is thus limited to sources observable from Effelsberg. The 6.7 GHz methanol masers are good tracers of the large scale structure of the magnetic field in the massive star formation regions at densities of nH2 ≈ 108 cm-3. However, here we have shown that previous magnetic field strength determinations were in error and that the exact Zeeman splitting coefficient is extremely uncertain. Assuming a calculation of the 6.7 GHz A-type methanol transition Landé g-factor extrapolated from laboratory measurements for an average of E-type methanol 25 GHz transitions, the geometry corrected total magnetic field in the methanol maser region  ⟨|B|⟩  = 120 mG. This is an order of magnitude higher than expected and likely indicates the g-factor cannot simply be extrapolated. Still, instrumental effect or non-Zeeman interpretations of the observed splitting are unlikely as no relation between maser flux and splitting is observed and as the observations are reproduced with a number of different telescopes and for different methanol maser transitions. The Zeeman splitting uncertainty has been shown to have little effect on the linear polarization measurements presented in other papers, but the field determination accuracy will only improve with new laboratory measurements of the correct g-factors. Still, we have detected significant Zeeman splitting in 76% of the sources with peak fluxes down to  ~20 Jy, which, unless the true g-factor is much more than an order of magnitude wrong, suggests relatively strong magnetic fields are widespread.


1

The 100-m telescope at Effelsberg is operated by the Max-Planck-Institut für Radioastronomie (MPIfR) on behalf of the Max-Planck-Gesellschaft (MPG).

Acknowledgments

W.V. thanks Alex Kraus for his help setting up the Effelsberg observations. W.V. also thanks Vladimir Strelnitski for pointing out the calculation error in the previous publications and Simon Ellingsen for communicating the results of his Hobart telescope Zeeman splitting confirmation. We further acknowledge useful discussion with Anuj Sarma and Vincent Fish.

This research was supported by the Deutsche Forschungsgemeinschaft (DFG) through the Emmy Noether Research grant VL 61/3-1.

References

Online material

thumbnail Fig. 5

Total intensity spectra (bottom) and Zeeman splitting (top) for all the sources of our sample with a significant detection. The Zeeman splitting is derived using the “running” cross-correlation method (see Paper I).

Open with DEXTER

thumbnail Fig. 6

Total intensity spectra of the sources in our sample observed with Effelsberg for which no significant Zeeman splitting was detected.

Open with DEXTER

thumbnail Fig. 7

Total intensity spectra of the sources in our sample observed with Parkes for which we were unable to determine the Zeeman splitting.

Open with DEXTER

Table 2

Parkes observations.

All Tables

Table 1

Zeeman splitting results.

Table 2

Parkes observations.

All Figures

thumbnail Fig. 1

Observed splitting between the right- and left-polarization spectra vs. the maser peak flux. The lack of an increase of the splitting with flux argues against a non-Zeeman interpretation of the observed splitting.

Open with DEXTER
In the text
thumbnail Fig. 2

Total intensity spectrum (bottom) and Zeeman splitting (top) for G111.43+076 (NGC 7538). The red solid line and square symbols denote the measurements presented in Paper I, taken at Nov. 12th 2007, the black dashed line and solid circles are the measurements presented here.

Open with DEXTER
In the text
thumbnail Fig. 3

Distribution of observed Zeeman splitting (thick solid line) in the Effelsberg 6.7 GHz methanol maser sample. Over-plotted are models for the line of sight Zeeman splitting in the case of where the magnetic field is taken to have a Gaussian distribution with a Zeeman splitting dispersion of σZ in each dimension. Following a simple analysis, the best fitted model has σZ = 0.4 m s-1 which would corresponds to a full 3-dimensional equivalent Zeeman splitting of  ⟨ ΔVz ⟩  ≈ 0.62 m s-1. As an indication, using the uncertain laboratory extrapolated g-factor (Sect. 3) this corresponds to  ⟨ |B| ⟩  ≈ 120 mG.

Open with DEXTER
In the text
thumbnail Fig. 4

Magnetic field direction derived from the methanol maser Zeeman splitting observations presented in this paper projected onto the Galactic plane. The symbols are the observed star forming region with kinematic distances from Pestalozzi et al. (2005) unless better distances were available. The open circles and crosses indicate a clockwise and counterclockwisemagnetic field direction respectively. The approximate location of the spiral arms is indicated as taken from Taylor & Cordes (1993).

Open with DEXTER
In the text
thumbnail Fig. 5

Total intensity spectra (bottom) and Zeeman splitting (top) for all the sources of our sample with a significant detection. The Zeeman splitting is derived using the “running” cross-correlation method (see Paper I).

Open with DEXTER
In the text
thumbnail Fig. 6

Total intensity spectra of the sources in our sample observed with Effelsberg for which no significant Zeeman splitting was detected.

Open with DEXTER
In the text
thumbnail Fig. 7

Total intensity spectra of the sources in our sample observed with Parkes for which we were unable to determine the Zeeman splitting.

Open with DEXTER
In the text

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.