VLBA imaging of a flaring 12.2 GHz methanol maser S. Goedhart 1 , V. Minier 2 , D.J. van der Walt 3 , M.J. Gaylard 1 e-mail: sharmila@hartrao.ac.za 1. Hartebeesthoek Radio Astronomy Observatory, PO Box 443, Krugersdorp 1740, South Africa 2. Dept. of Astrophysics & Optics, School of Physics, University of New South Wales, NSW, 2052, Australia 3. Space Research Unit, Physics Department, Potchefstroom University for CHE, Private Bag X6001, Potchefstroom 2520, South Africa Abstract The class II methanol maser G9.62+0.20E has been found to undergo periodic flares at both 6.7- and 12-GHz. Single-dish observations at HartRAO enable us to monitor the changing intensity of the maser with high time resolution. However, high angular resolution imaging is necessary in order to understand the underlying mechanism causing the flares. The flare peaking in October 2001 was observed during a series of seven VLBA observations spaced over a period of three months. The results of this monitoring program are presented here. Introduction Class II methanol masers appear to be closely associated with newly formed massive stars (Menten 2002). These stars are still deeply embedded in their natal molecular cloud and thus cannot be observed directly. Hence the methanol masers are a useful tool for probing conditions in these regions. The massive star formation region G9.62+0.20 contains a number of HII regions, indicating the presence of massive young stars at different stages of their (proto)stellar evolution (Hofner et al. 1996). High resolution imaging has shown that the most powerful known 6.7 GHz methanol maser (Phillips et al. 1998) as well as 12.2 GHz methanol masers (Minier et al. 2000) are associated with a compact radio source which may be an ultracompact (UC) HII region (component E in Garay et al. 1993) or an ionized stellar wind around a young B0 star (Hofner et al. 1996). G9.62+0.20E has been monitored at HartRAO at 6.7 GHz since January 1999 and at 12.2 GHz since January 2000. A series of flares with a period of 246 days has been observed at both frequencies (Goedhart et al. 2003). Contour plots of the time series observed at HartRAO over the last four years are shown in Figures 1 and 2. While the maser is more powerful at 6.7 GHz, its variability is far greater at 12.2 GHz. During the course of a flare the intensity at 12.2 GHz can double within two weeks. The 12.2 GHz observations at HartRAO are not sensitive enough to monitor the weaker features in this source, but a time lag of 20 days between the 6.7 GHz features peaking at 1.21 and -0.14 km.s -1 can be seen. The velocity structure of the region is quite complex, with a number of spatially distinct maser spots at the same velocity, as well as spots with different velocities at the same spatial location (Minier et al. 2000). As a result of this complexity, it is not possible to distinguish between variability in the various maser spots using spectroscopy alone. The indication of the existence of a time lag in the progress of the flare is also intriguing. The sequence of the flare through the maser spots can help shed light on the origins of the flare. Figure 2: HartRAO time series for the 6.7 GHz methanol maser. Figure 1:HartRAO time series for the 12.2 GHz methanol maser. Observations The observations were done using the full VLBA at 12.2 GHz with both left and right circular polarisations. The continuum source J1733-1304 was used as a calibrator. The total observing time per epoch was 6 hours. A bandwidth of 1 MHz was used, giving a velocity resolution of 0.05 km.s -1 for 512 channels. Figure 3 shows the timing of the observations relative to the flare at the dominant maser peak. The data were reduced using standard procedures in AIPS. The amplitude calibration was done using a reference spectrum with minimal noise and RFI in the task ACFIT. In order to maintain a consistent beam pattern, antennas which were off-line during any epoch were completely flagged out of the final data set for all epochs before the data cubes were generated with IMAGR. Results: Sixteen maser regions, with distinct spatial positions and velocity ranges, were identified. Figure 4 * shows the distribution of the masers. The field can be loosely divided into three regions: Features A, B, C1, C2 and D are in the top left quadrant. Features E1--5, F1, F2 and G2 are in the bottom left quadrant. Features G1, H, and I are in the bottom right quadrant. The zero-moment images for each epoch are shown in Figure 5 * . There do not appear to be any changes to the morphology of the spots during the course of the flare. No new spots are formed and the spots do not appear to move relative to the phase center. The only visible effect is an increase in spot brightness and a subsequent dimming. Slight variations in spot size and orientation are due to minor variations in the beam pattern. Zero-moment maps were generated for each velocity range and epoch. The total flux for each region was found using BLSUM after the noise in the maps had been BLANKed out. The noise was estimated using IMSTAT. Errors in the fluxes are typically one percent of the total flux. Figure 6 shows the time-series found for the masers. The time series are grouped by region in the plot. Three different trends in the flare pattern are apparent. The observations were timed to capture the flare in the dominant features at A, B, C and D. As a result the flare is well sampled for this region. However, the flare appears to peak slightly earlier in the other two regions. The time resolution is not good enough to determine the exact time delay, but the progress of the flare does appear to be dependent on the position of the masers. References: Garay G., Rodriguez L.F., Moran J.M., Churchwell E., 1993, ApJ, 418, 368 Goedhart S., Gaylard M.J., van der Walt D.J., 2003, MNRAS, 339, L33 Gooch, R., 1996, PASA, 14, 106 Hofner P., Kurtz S., Churchwell E., Walmsley C.M., Cesaroni R., 1996, ApJ, 460, 359 Menten K., 2002, in V. Migenes & M.J. Reid (eds.) Cosmic Masers: From Protostars to Black holes, Proceedings of IAU Symposium 206, ASP Conference Series, pp 125-126 Minier V., Booth R.S., Conway, J.E., 2000, A&A, 362, 1093 Phillips C.J., Norris R., Ellingsen S.P., McCulloch P.M., 1998, MNRAS, 300, 1131 Figure 3. Timing of the VLBA observations relative to the flare cycle. The time series for the peak velocity channels at 16.7- and 12,2 GHz are folded modulo 246 days. Figure 4. Distribution of the maser features in G9.62+0.20E. The figure is generated from the fourth VLBA observation, when features A, B, C and D were close to their maximum intensity. The colour-coded contour levels are at 0.2 Jy/beam in each velocity channel. The inset shows the total power spectrum. The colour bars superimposed on the spectrum indicate the corresponding colour range in the contour map. The features are labeled on the contour map and on the spectrum. Conclusion A flare in the 12.2 GHz methanol maser G9.62+0.20E was imaged in a series of seven observations with the VLBA. We have determined that no new masers are excited during the flare, and that there is no change in the morphology or relative positions of the existing masers. Therefore the cause of the flare does not originate in the masing region itself. The flare could be caused by an increase in the pumping photons, or the seed photons. It is believed that the masers are pumped by infrared photons, and continuum radiation from a background HII region is being amplified. Observations in the infrared and radio continuum will be needed to confirm the origin of the flare. Figure 5. Zero-moment images for each epoch. Figure 6. VLBA time series for the different maser features in G9.62 +0.20 E. The three panels are divided according to the spatial distribution of the masers. * Spot maps generated using the KARMA package (Gooch 1996)