The 3-4 Nov 1993 storm in the context of recurrent geomagnetic 
activity during the decline of solar cycle 22

N. U. Crooker
Center for Space Physics, Boston University

	Continuous solar wind data and the x-ray and white light 
views of the sun available during the decline of solar cycle 20 in 
1973-1975 led to major advances in solar-terrestrial physics.  
Coronal holes were discovered as the source of recurrent high-
speed streams, and coronal mass ejections (CMEs) were discovered 
as the source of transient interplanetary disturbances.  
Thereafter coronal holes and CMEs became the accepted sources of 
recurrent and nonrecurrent geomagnetic storms, respectively.  
Recent research on the 1974 data, however, has resulted in a 
broader view of the source of recurrent storms.  Crooker and 
Cliver [1994] re-emphasized the importance of an older association 
between recurrent storms and corotating interaction regions 
(CIRs), raised the possibility that CMEs may play a central role 
in recurrent as well as nonrecurrent storms, and demonstrated that 
the Russell-McPherron effect exerts strong control over recurrent 
activity [Russell and McPherron, 1973].  Although continuous solar 
wind data are not available for most of the decline of cycle 22, 
some of these effects are apparent in 27-day recurrence plots of 
the (provisional) Dst index of geomagnetic activity in Figure 1 
and Figure 2, upon which the inferred interplanetary sector 
structure has been superposed.
	Figure 1 shows the period from June solstice through 
September equinox to December solstice in 1993, ordered by Bartels 
rotation.  The superposed shading indicating IMF polarity was 
obtained from the OMNI spacecraft data set, when available, and 
supplemented with polarities from the Stanford mean solar magnetic 
field recurrence maps published in Solar-Geophysical Data.  The 
resulting sector structure agrees well with the polarity maps 
provided by S.Watari for this period, based on interpolated 
spacecraft measurements alone.  In particular, it shows the 
evolution from a four-sector to a two-sector pattern pointed out 
by Watari.
	The 3-4 Nov 1993 storm began at the end of Bartels rotation 
2188.  As pointed out by D. Knipp, it was the fourth in a series 
of five recurrent storms, with Dst maxima of -78, -169, -82, -116, 
and -111 nT, respectively.  (Unfortunately, these storms do not 
stand out in the figure because they span days from the end of one 
rotation to the beginning of the next.)  This was the only series 
of recurrent storms in the interval.  The storms were typical of 
the 1974-5 recurrent storms.  They occurred in the sector with IMF 
polarity favorable for the Russell-McPherron effect for that 
season--away polarity, in this case, for the September equinoctial 
period.  The peak of each storm occurred shortly after sector 
boundary passage, where the combined effects of CIRs and CMEs can 
create strong southward magnetic fields.  As pointed out in the 
background information for the 3-4 Nov event, the available Yohkoh 
and Mauna Loa data show activity on the sun consistent with the 
view that CMEs (and/or less spectacular but, nevertheless, 
transient features) were responsible for peak activity in these 
recurrent storms.  Each of the five storm peaks was followed by 
sustained, low-level activity that extended across the sector 
passage.  Sustained activity has been ascribed to the combination 
of the high speed of coronal hole flow and southward IMF created 
by its intrinsic Alfvenic fluctuations [Tsurutani and Gonzalez, 
1987].  It is clear from the 1974 observations, however, that 
sustained activity is strongly controlled by the Russell-McPherron 
effect, which provides an added southward IMF from the projection 
of the predominantly ecliptic magnetic fields onto Earth's dipole-
ordered magnetospheric (GSM) coordinate system.  This effect is 
also apparent in Figure 1.  For example, the storm that followed 
the 3-4 Nov storm, on 18-19 Nov, occurred in a toward sector and 
had no sustained activity.  In general, Dst hugs the baseline in 
the shaded regions of toward polarity in the figure and hovers 
below the baseline in the open regions of away polarity.
	Figure 2 shows the period from December, 1993, solstice 
through March equinox to June, 1994, solstice.  Although this 
period goes beyond the storm interval of interest, it is included 
to illustrate similar recurrence characteristics under a stable, 
two-sector configuration, as in 1974.  In the solstitial rotations 
2191 and 2197, at the top and bottom, there was no difference 
between activity in the two sectors; but in the five intervening 
rotations, Dst was clearly more depressed in the toward compared 
to the away sector, opposite to Figure 1, consistent with the 
Russell-McPherron effect for this March equinoctial season.  That 
there was any sustained activity at all in the away sector 
suggests that a baseline adjustment is needed for these 
provisional Dst values.  Peak activity at the leading edge of the 
toward sectors was relatively weak compared to the sustained 
activity, suggesting weak speed gradients in the associated CIRs 
and/or no CME activity associated with that sector boundary.  The 
only two pronounced storms, peaking at the beginnings of rotations 
2193 and 2195, occurred at the other sector boundary, where they 
fell into the same recurrence pattern as the five storms in Figure 
1.  These form the subject of two coordinated studies undertaken 
by STEP and clearly were produced by CMEs.  In contrast to Figure 
1, however, the active streamer belt region coincided with passage 
from favored to unfavored polarity, which places the storms at the 
beginning of the unfavored sector.  Thus no following sustained 
activity would be predicted.  What appears to be sustained 
activity may be the result of the baseline problem.
	In summary, the 3-4 Nov 1993 event was a typical recurrent 
storm.  Peak activity occurred shortly after passage into the 
sector favored by the Russell-McPherron effect for that season, 
and it was followed by sustained activity throughout most of the 
remaining sector passage.  The peak activity most likely was 
caused by southward IMF from streamer belt transients, possibly a 
CME, which may have been enhanced both by compression in a CIR and 
by the Russell-McPherron effect.  The sustained activity can be 
more directly ascribed to high speed flow from a coronal hole, 
consistent with the 1970's view of recurrent activity; but its 
amplitude was strongly controlled by the Russell-McPherron effect.

References
Crooker, N. U., and E. W. Cliver, Postmodern view of M-regions, J. 
Geophys. Res., 99, 23,383-23,390, 1994.
Russell, C. T., and R. L. McPherron, Semiannual variation of 
geomagnetic activity, J. Geophys. Res., 78, 92-108, 1973.
Tsurutani, B. T., and W. D. Gonzalez, The cause of high-intensity 
long-duration continuous AE activity (HILDCAAS):  Interplanetary 
Alfven wave trains, Planet. Space Sci., 35, 405-412, 1987.

	Acknowledgments.  The provisional Dst index values were 
obtained from NOAA, World Data Center A for Solar-Terrestrial 
Physics, with the special help of H. Coffey, and the OMNI 
spacecraft data were obtained from the National Space Science Data 
Center, both through the World Wide Web.  This research was 
supported by the National Science Foundation under grant ATM94-
21814.