Magnetic Fields in the Gas Feeding the Galactic Centre

A magnetic field near the centre of our Galaxy might be very important for gas dynamics - in particular, it could help channel material falling onto the (candidate) black hole at the centre, Sgr A*. There is good evidence for strong organised magnetic fields, both from Zeeman data (Plante, Lo & Crutcher 1995) and dust polarization (Hildebrand et al. 1993, Novak et al. 2000, and recent SCUBA polarimetry by Chrysostomou et al.). However, the view towards the inner few parsecs of our Galaxy is very complicated - so these techniques might mix up different magnetic field components from any of the clouds along the 8.5 kpc line of sight!

Our solution to this was to use spectral line polarimetry, and thus to separate out all the foreground clouds by velocity. Small polarizations arise in molecular lines because population imbalances among the magnetically-split rotational levels can produce different emission in the planes perpendicular or parallel to the magnetic field (Goldreich & Kylafis 1981). Spectro-polarimetry has been offered at the JCMT for several years, but always as a non-standard observing mode, and it is only now that the data reduction techniques have become streamlined and easy to use.

Figure 1: CO 2-1 integrated intensity toward the Galactic Centre.

These results are from our observations of the Galactic Centre in CO J=2-1. Figure 1 shows one of the main gas streamers feeding into the Galactic Centre, at an LSR velocity of +80 km/s. This streamer runs from the north end of the circumnuclear `2 pc ring' towards Sgr A* at the centre of the map. Figure 2 shows the polarization results for a point at the top of this streamer (dRA,dDec = 0,+75"): the red line is the total intensity CO spectrum, the green line is the polarization percentage spectrum (multiplied by 10 for clarity), and the blue line is the polarization direction (in degrees). Although there are some variations among the positive velocity features, the polarization direction in the streamer gas (+80 km/s) is quite similar to the other clouds, for example the '2 pc ring' which has a velocity at this point of about +65 km/s. The net direction is about -10 degrees, and is perpendicular to dust polarization (+77 degrees: Hildebrand et al. 1993). The only dramatically different result is for a second streamer at -20 km/s, where the polarization direction diverges by about 60 degrees.

Figure 2: Polarization as a function of velocity at one position toward the Galactic Center (dRA,dDec = 0,+75"). Red line: CO spectrum; green line: polarization percentage spectrum (multiplied by 10 for clarity); blue line: polarization direction (in degrees).

Thus there seems to be a largely uniform magnetic field direction for the gas clouds orbiting around and infalling onto Sgr A*. Also, the inferred magnetic field direction is close to north-south - the configuration required to help material fall onto the black hole. In fact, our full data set of four positions in the north-south streamer suggests slight curvature of the magnetic field, roughly following the ionized gas 'mini-spiral' (Jackson et al. 1993). Spectro-polarimetry has thus given us more robust evidence for this magnetic configuration than did previous observing techniques - and has many other potential applications for gas clouds with velocity structure. The 'ROVER' project (http://www.jach.hawaii.edu/~jsg/rover/rover.html) aims to bring a new spectro-polarimetry upgrade to the JCMT in about 2 years time.

Goldreich P. & Kylafis N.D., 1981, ApJ 243, L75

Hildebrand R.H. et al., 1993, ApJ 417, 565

Jackson J.M. et al., 1993, ApJ 402, 173

Novak G. et al., 2000, ApJ 529, 241

Plante R.L., Lo K.Y. & Crutcher R.M., 1995, ApJ 445, L113