Proceedings of the Workshop
"The Magellanic Clouds and Other Dwarf Galaxies"
of the Bonn/Bochum-Graduiertenkolleg

Comparison of H I and magnetic field structures

in our Galaxy

H.-R. Klöckner, P.M.W. Kalberla, and U. Klein

Radioastronomical Institute, University of Bonn

Received 12th May 1998
Abstract. The Leiden/Dwingeloo 21 cm line survey (Hartmann & Burton 1997) was used to search for a correlation between the H I column density distribution and magnetic fields. The structure of the local galactic magnetic field was estimated from the polarization surveys of Mathewson & Ford (1970), the rotation measures of extragalactic radio sources of Tabara & Inoue (1980), Simard-Normandin et al. (1981), Broten et al. (1988), and from pulsar measurements by Taylor et al. (1993).

No general correlation could be detected between H I column densities and rotation measures. A number of filamentary H I features were found which correlate well with stellar polarization data of Mathewson & Ford (1970). These filaments seem to be located within the local interstellar medium (d ≤ 1 kpc).


1. Introduction

Interstellar dust, gas, magnetic fields, and cosmic rays are expected to be interwoven and closely related to each other (Parker 1969). H I, although neutral, is expected to be closely tied to interstellar magnetic fields because even a low-level ionization of ne ∼ 4·10-4 cm-3 is sufficient to couple the gas to the magnetic fields. Also grains may be considered to be tied to the magnetic field because electric charges on their surfaces are caused by the photoelectric effect. Visualizations of the Leiden/Dwingeloo 21 cm line survey (LDS, Hartmann & Burton 1997) show a rich wealth of filamentary features, which are highly suggestive of beeing associated with magnetic field structures. Since the accuracy of the LDS is an order of magnitude better than previous H I surveys, we decided to search for possible associations between gas, dust and magnetic fields.

2. Rotation Measures and H I

Rotations measures (Tabara & Inoue 1980; Simard-Normandin et al. 1981; Broten et al. 1988; Taylor et al. 1993) have been compared with the H I column density distribution. We found no global correlation between the rotation measures and H I column densities. The distribution of rotation measures shows a patchy structure, which is also not visible in the H I data. We conclude that there is either no correlation between neutral and ionized components in the interstellar medium or the magnetic field is dominated by turbulence and field reversals.

3. Stellar Polarization and H I

Correlating stellar polarization data (Mathewson & Ford 1970) with LDS H I maps, a large number of filaments can be found which are highly suggestive of a possible correlation. In summary, the results are:

For H I gas at high- or intermediate velocities (|vLSR| > 11 km s-1), there is no significant correlation with the polarization data. Most of the stellar data are from distances of d < 1 kpc. High-velocity clouds are assumed to be at larger distances, therefore no correlation is to be expected from such observations. For intermediate-velocity clouds the distances are less certain. Our result is then consistent with the assumption that intermediate-velocity clouds do not belong to the local interstellar medium.

Channel maps at constant velocities may be used to study details of the correlation between H I gas and stellar polarization. Individual H I filaments which are associated with optical polarization have a typical velocity range of 5 to 10 km s-1. H I gas with such line widths indicate temperatures of few hundred Kelvin. We compared the filaments which are associated with stellar polarization with far-infrad emission, e.g. at 240 µm (derived from COBE maps, Hauser et al.) and found a less significant correlation from these data. This indicates that the H I filaments are most probably warm clouds (T > 30 K).

It is well known that radio loops I, III and IV are associated with H I filaments (Berkhuijsen 1971). In Fig. 1 the radio loops are indicated. Loops I and IV are clearly associated with H I filaments as well as with stellar polarization. We find no obvious correlation with loop III.

[Click here to see Fig. 1!]

References


Links (back/forward) to:

First version: 04thAugust,1998
Last update: 29thSeptember,1998

Jochen M. Braun   &   Tom Richtler
 (E-Mail: jbraun|richtler@astro.uni-bonn.de)