The arrival of the interplanetary signature of the 6 January 1997 Halo coronal mass ejection (CME) at the Earth on January 10 has been successfully predicted by assuming a constant propagation velocity of 450 km/s (LASCO press release).
By itself, the appearance of a Halo event at the Sun can not be used to predict geomagnetic activity. Only those interplanetary CME counterparts (specifically, magnetic clouds) that have a magnetic field direction anti-parallel to the magnetic field of the Earth react strongly with the Earth's magnetosphere. To predict whether or not a CME will be geoeffective, we need information about the internal magnetic field of the CME.
We have shown (Zhao and Hoeksema, 1997) that one major factor that determines whether or not a CME will generate a long duration southward interplanetary magnetic field (IMF) at 1 AU (the Bs event) is the orientation of the central axial magnetic field in the disappearing filament (DSF) associated with the CME. The intensity and duration of the associated Bs event become increasingly stronger and longer as the DSF central axial field direction turns from north to south.
This poster shows how we determined the orientation (specifically, the heliographic latitude) of the central axial magnetic field in the DSF associated with the Jan. 6 CME (section 2) and how to infer the ecliptic latitude of the associated magnetic cloud (a cylindrical rope) (section 3) and the intensity and duration of the associated Bs event (section 4). The poster concludes with a discussion.
The SHINE Report (Webb, 1997) has indicated that the Jan. 6 CME was associated with a DSF observed between 13:01 and 14:53 UT Jan. 6 and centered at S24W01 (see the 17GHz images at 06:45 and 24:45 showing location of the DSF.
The magnetic fields in solar filaments have been found to be preferentially sinistral heliform in the northern hemisphere and dextral in the southern hemisphere (Martin et al., 1994). This segregation of filament field helicity by hemisphere has been confirmed by the field configuration observed in magnetic clouds (Rust, 1994; Marubashi, 1996).
To determine the central axial field direction of the Jan. 6 DSF with a dextral heliform, we need the field polarity distribution around the DSF at the time when the quiet filament disappeared. The high-cadence (every 96 minutes) SOI-MDI observations of the photospheric magnetic field make this possible. Figure 1 displays 14 panels with the observation times marked at the upper left of each panel and the maximum and minimum values of the observed line-of-sight field at upper middle and right. The polarity inversion lines in the panels are located where the polarity changes from positive (pink) to negative (light blue). Figure 1 is obtained by lowering the spatial resolution of the SOI-MDI magnetic images. The NW to SE trending polarity inversion line just south of disk center around midday is nearly parallel to the DSF. Using Martin's handedness of filament field rotation and the polarity distribution of the photospheric magnetic field around the DSF we are able to determine that the heliographic latitude of the central axial field direction of the Jan. 6 DSF is 40S degrees with a standard deviation of 10 degrees.
We have established a linear regression relation for the central axial field directions of the interplanetary magnetic clouds and the associated solar filaments. It shows that the central axial field direction changes slightly while the magnetized plasma ejector propagates into the interplanetary medium. Using Figure 2 we infer that the ecliptic latitude of the Jan. 10 cloud central axial field direction is about 32S degrees. The familiar view is that magnetic rope axes determined for magnetic clouds lie in the ecliptic plane, perpendicular to the Earth-Sun line (e.g., Lepping et al., 1990). Based on 30 magnetic clouds, we have shown (Zhao and Hoeksema, 1997) that magnetic field central axial field directions are evenly distributed between ecliptic latitudes of 90S and 90N degrees, and their occurrence frequency peaks around east and west.
We have defined a magnetic cloud Bs event as the interval of southward IMF observed by a spacecraft as a magnetic cloud is passing by the spacecraft (Zhao and Hoeksema, 1997) and found that the intensity (the maximum Bs field) and duration of magnetic cloud Bs events depend on the direction and strength of the magnetic cloud central axial field, rather than the azimuthal field (Zhao and Hoeksema, 1997). Based on the ecliptic latitude of the Jan. 6 magnetic cloud central axial field, as inferred from the heliolatitude of the DSF central axial field direction, and the dependence of magnetic cloud Bs events on the cloud central axial field direction (see Figure 3 ), the intensity and duration of the magnetic cloud Bs event are inferred to be about -17 nT and 16 hours, respectively, with standard deviations of 5 nT and 4 hours. The intensity and duration observed by WIND is about 15 nT and 13 hours (see The Jan. 10 WIND Survey).
Based on the observations of the Jan. 6 DSF that is supposed to be associated with the Jan. 6 Halo CME propagating toward the Earth and the photospheric magnetic field polarity distribution inferred from the SOI/MDI magnetic images observed near the time when the DSF took place, we are able to determine the directions of the central axial field in the DSF and the associated magnetic cloud, and predict the intensity and duration of the magnetic cloud Bs event. The predicted intensity and duration of the Bs event quite well agrees with that observed.
Many Bs events that generate magnetospheric storms occur both in magnetic clouds and in shock sheathes between the shock and driver gas. Given the CME speed, the possibility for an associated shock to be generated may be estimated.
The shock sheath Bs event is caused by compression and draping of preexisting southward magnetic field lines at the leading edge of driver gas. The preexisting southward field is usually the ambient southward IMF, though it may also include a portion of the internal field of slow ICMEs (Burlaga et al., 1987; Zhao, 1992). Based on the Russell-McPherron effect, the presence or absence of the ambient Bs may be predicted using IMF polarity extrapolated from the observed photospheric field. The shock sheath Bs event may be precluded with a high possibility if there is no ambient Bs, whatever the CME speed (see Zhao et al., 1993 for the details).
More events are needed to further test our prediction scheme of magnetic cloud Bs events. The scheme is established on the basis of two sets of data and magnetic cloud models (Lepping et al., 1990; Marubashi, 1996). The scheme may be improved using one improved cloud model and more events.
Burlaga, L. F., K. W. Behannon, and L.W. Klein, Compound streams, magnetic clouds, and major geomagnetic storms, J. Geophys. Res., 92, 5725, 1987.
Lepping, R. P., J. A. Jones, and L. F. Burlaga, Magnetic cloud structure of interplanetary magnetic clouds at 1 AU, J. Geophys. Res., 95, 11957--11965, 1990.
Marubashi, K., Interplanetary flux ropes and solar filamanets, in AGU Monogram on Coronal Mass Ejections, in press, 1996.
Martin, S. F., R. Bilimoria, and P. W. Tracadas, Magnetic field configurations basic to filament channels and filaments, in Solar Surface Magnetism, edited by R. J. Rutten and C. J. Schrijve r, pp. 308--338, Kluwer, 1994.
Rust, D. M., Spawning and shedding helical magnetic fields in the solar atmosphere, Geophys. Res. Lett., 21 241-244, 1994.
Zhao, X. P., Interaction of fast steady flow with slow transient flow: A new cause of shock pair and interplanetary Bz event, J. Geophys. Res., 97, 15051--15055, 1992.
Zhao, X. P., J. T. Hoeksema, J. T. Gosling, and J. L. Phillips, Statistics of IMF Bz events, in Solar-Terrestrial Predictions Workshop--IV, edited by J. Hruska et al., pp. 712--719, Natl. Oceanic and Atmospheric Administration, Boulder, Vol. 2, 1993.