Photoionization of the Magellanic Stream

The Stream lies along a great arc which extends for more than 100 (e.g., Mathewson, Cleary & Murray 1974). Fig. 1 illustrates the relationship of the LMC to the Magellanic Stream above the Galactic disk (Mathewson & Ford 1984). We shall make the assumption that the Stream lies along a circular orbit, close to the X-Z plane, originating from the Lagrangian point between the LMC and SMC. The Cepheid distance moduli indicate that for the LMC , which implies a distance of 49.43.4 kpc (Feast & Walker 1987); for the SMC, which implies 58.34.0 kpc (Feast 1988). Thus, we shall assume an average galactocentric radius of 55 kpc for the Stream. This is an oversimplification since most computed orbits for the LMC-SMC system imply substantial ellipticity with the Galaxy at a focal point (e.g., Lin, Jones & Klemola 1995). Our model is consistent with the distance measured by Gardiner et al. (1994) towards MS VI, but not with the much smaller value of 20 kpc determined by Moore & Davis (1994).

 
Figure 2: Meridional plot showing the probable contribution of the LMC to the opaque-disk halo radiation field (solid lines). The dotted lines are for the opaque-disk model in Fig. 3. The position of the LMC in Galactic coordinates lies within 2 kpc of the plane Y=0 (Fujimoto & Sofue 1976). The figure shows a 100 kpc 100 kpc intersection of the non-axisymmetric radiation field in the plane Y=0. The dots represent the HI warp in the outer parts of the Galaxy close to the line of longitude (Burton 1988).

 
Figure 3: The predicted emission measure along the Magellanic Stream as a function of . The vertical axis has units of log(cm pc); these can be converted to log(Rayleighs) by subtracting 0.48. The dotted curves (top) assume an optically thin Galactic disk with and without the LMC ionizing field. The solid lines assume an opaque ionizing disk with (thin line) and without (thick line) a bremsstrahlung halo. The LMC contribution to the opaque disk(+halo) model is shown by the short-dash () and long-dash () curves. The dot-dash curve is () predicted from the upper side of the Magellanic Stream due to the bremsstrahlung halo. The solid points are the measurements of Weiner & Williams (1996).

In Fig. 2, we present a meridional plot of the halo radiation field for . While the distance to the Magellanic Stream is uncertain, the expected emission measure for the opaque disk model should be easily detectable. For distances of (20,40,60) kpc, takes values of (710,215,105) phot cm s (Fig. 3). From equation (7), the expected values are (9.0,2.7,1.3) cm pc, or equivalently, (18,5.4,2.6) erg cm s arcsec. The Weiner & Williams (1996) detections along the stream are 370, 210 and 200 milliRayleighs or, equivalently, values of (1.1,0.63,0.60) cm pc. The measurements of Weiner & Williams (1996) are within range of the model values, particularly since the Stream distance is at the far end of our range.

In Fig. 3, we present the predicted emission measure along the Stream after projecting the clouds into the X-Z plane, where the observer is assumed to be at the Galactic Centre. If we assume is close to unity and remains constant along the Stream, several conclusions follow immediately. The Galactic disk is unlikely to be transparent to ionizing photons otherwise the Magellanic Stream would be mostly ionized. The shape of the curve gives an independent assessment of the disk opacity, but this is sensitive to departures from a circular trajectory. With relatively few unknowns, the mean UV opacity of the Galactic disk can be determined after a comprehensive observational campaign along the Stream. If the Stream orbit is highly flattened (Moore & Davis 1994), the solid line in Fig. 3 becomes significantly more boxy at large , and possibly even sharply rising towards the edges before turning over. The expected value of at MS VI () could be almost an order of magnitude higher for a distance of 20 kpc compared with our adopted value. The major limitation of our model is the poorly known cloud geometry and HI covering fraction.

In the interests of brevity, we do not discuss alternative ionizing sources (e.g. shock or halo sources) as these are expected to be entirely negligible. For illustrative purposes only, we include the expected ionization from the LMC and halo bremsstrahlung in Fig. 3. For the coronal gas, we assume an isothermal sphere with central density cm, scale length 10 kpc and electron temperature K (0.2 keV). The LMC is treated as a point source radiating ionizing photons per second. For a complete discussion, we refer readers to Bland-Hawthorn & Maloney (1996).

The influence of the corona is only likely to be observable at extreme angles where emission from the upper cloud face is expected to dominate. At angles larger than 150, the isothermal halo acts much like a distant point source so would be difficult to distinguish from the LMC ionization. The LMC radiation field is not expected to substantially ionize the Magellanic Stream (MS I-VI) although, presumably, it has a major impact on the outer parts of the Milky Way in the direction l=270 (see Figs. 2 and 3). If there are no UV-bright companions, the outer extremities of opaque disks fall inside a `toroidal shadow' which sees only a very weak ionizing field from the Galactic halo. If the outer warp in the HI disk is not severe ( from center to edge), the ionization of cold gas at large radius should be dominated by the cosmic UV background. The current 2 upper limit on the flux, =3.8 (q.v. Bland-Hawthorn 1997), indicates that the cosmic background is expected to produce an equivalent emission measure less than = 0.05 cm pc.

In summary, for a mean Stream distance of 55 kpc, if , the detections indicate perpendicular to the Galactic disk such that only 5% of the ionizing radiation from the disk escapes into the halo. Notably, Domgorgen & Mathis (1994) have obtained the same result using an entirely different approach. While OB stars should dominate the ionization balance, just how the ionizing radiation escapes from the star-forming regions into the halo is still somewhat unclear, although recent theoretical models have begun to address this issue (Miller & Cox 1993; Dove & Shull 1994).

© Copyright Astronomical Society of Australia 1997