The interaction of impulsive solar wind discontinuities with the magnetosphere: a multi-satellite case study

Planer. Spme Sk. Vol. 3X. No. 7. pp 841-850. 1990 Pnnted m Great Bntain.

0032-ofJ33/90 %3.00+0.00 <.” 1990 Pergamon Press plc

THE INTERACTION OF IMPULSIVE SOLAR WIND DISCONTINUITIES WITH THE MAGNETOSPHERE :

A MULTI-SATELLITE CASE STUDY

J. LABELLE” Department of Physics and Astronomy, Dartmouth College, Hanover. NH 03755, U.S.A.

L. M. KISTLER, R. A. TRE~MANN and W. BA~MJOHANN Max-Planck Institut fur extraterrestrische Physik, D-8046 Garching, F.R.G.

D. G. SIBECK Applied Physics Laboratory, Johns Hopkins University, Laurel, MD 20707. U.S.A.

D. N. BAKER Laboratory for Extraterrestrial Physics, Goddard Space Flight Center, Greenbelt, MD 20771, U.S.A.

and

R. D. BELIAN Los Alamos National Laboratory, Los Alamos, NM 87545, U.S.A.

jReeeit>ed 12 February 1990)

AIrstract-During the magnetic storm of 4-5 September 1984. the outbound ~~PTE~IR~ spacecraft stayed just outside the expanding Earth’s bow shock for a period of 7.5 h. During this interval the solar wind ram pressure and the interplanetary magnetic field remained approximately constant, except for two distinct impulsive pressure pulses lasting a few minutes each. These pulses coincided with discontinuities in which the IMF changed direction by about 90’. and the solar wind kinetic pressure decreased slightly. Accompanying these discontinuities. the following magnetospheric signatures were observed : (1) immedi- ately after each discontinuity. the growth phase of a substorm commenced as evidenced by decreases in the Rux of energetic ions at geosynchronous orbit on both the dayside and nightside, and these were followed by particle injection events on the nightside 20-35 min after the discontinuities; (2) equatorial magnetograms recorded sudden impulses in the magnetic field : and (3) after a delay of 8-12 min. the IRM detected outward motion of the bow shock, and in each case about an hour passed before the outward- moving satellite caught up again with the shock. Overall, the magnetospheric and ground signatures of the first discontinuity were larger than those of the second. It is speculated that the IMF direction and other factors such as the local time and latitude of the around stations may be important factors in determining the magnitude of these signatures.

INTRODUCTlON

Discontinuities in the solar wind kinetic pressure and magnetic field affect the position of the magnetopause and initiate geomagnetic activity. For example. south- ward turning of the interplanetary magnetic field (IMF) is associated with the onset of reconnection at the dayside, consequent erosion of the dayside mag- netopause. enhanced transfer of magnetic flux to the magnetotail. and release of a portion of this magnetic energy in the form of drifting energetic particles in the magnetosphere, resulting in an associated signature in ground magnetograms. Northward turning of the IMF after prolonged periods of southward IMF also seems to be associated with substorm expansive

*Also at: Max-Pianck Institut fur extraterrestrtsche Recently there has been a resurgence of interest Physik. Garching. F.R.G. in the magnetospheric effects of very short-lived

phase effects (Rostoker, 1983 ; and references therein).

Changes in the solar wind kinetic pressure move the magnetopause inward or outward, thus affecting the tlux of energetic particles observed at any particular location within the magnetosphere and leading to increases or decreases in the H-components of low- latitude ground magnetograms (e.g. Chap. 4 of Akasofu. 1977; Nishida, 1978). The Earth’s bow shock should of course in general follow the mag- netopause in such inward/outward excursions. It has been predicted that the shock might under some con- ditions move relative to the magnetopause in response to interplanetary discontinuities (Auer. 1974; Vijlk and Auer. 1974; Neubauer, 1975), although direct observational evidence for this effect has not been published.

841

842 J. LABELLE et al.

(- 1 min) impulsive solar wind dynamic pressure

pulses or IMF discontinuities (e.g. Friis-Christensen

et al.. 1988; Fairfield et al., 1990; Sibeck et al., 1989a,b; Potemra et al.. 1989). In particular, it has been specu- lated that such impulsive phenomena might generate

signatures in satellite plasma and magnetic field data

similar to those expected from flux-transfer-events or other transient phenomena such as magnetic “holes” (Sibeck et al., 1989b). Furthermore, hot diamagnetic

cavities (“funnies”), observed in the solar wind

upstream of the bow shock (Thomsen et al., 1986;

Schwartz et al., 1985), may be manifestations of a bow

shock-IMF discontinuity interaction (Paschmann et

al., 1988 ; Schwartz et al., 1988 ; Thomsen et al., 1988;

simulation by Burgess and Schwartz, 1988). In this

study, we extend the previous work on such inter- actions by examining a case in which the bow shock

is observed to respond to a pair of discrete dis- continuities in the solar wind dynamic pressure and IMF orientation. We show simultaneous observations

from inside the magnetosphere at geosynchronous orbit and from low-latitude magnetic observatories.

The observations indicate : (1) the bow shock moves inward promptly as expected due to enhancements in the solar wind dynamic pressure ; (2) the fluxes of the energetic particles at geosynchronous orbit display a

repeatable variation, first decreasing then increasing; and (3) sudden increases in the Earth’s magnetic field

measured on the ground occur which depend on IMF orientation as well as the latitude and longitude of the

ground observing station.

INSTRUMENTATION

The AMPTEjIRM satellite was launched in August

1984. into a highly elliptical orbit (apogee of about 18.7 Earth radii) which brought it often into the neighborhood of the bow shock for long time periods. The satellite instrumentation included a flux-gate vec-

tor magnetometer accurate to approx. 0.1 nT (Liihr et al., 1985) and a plasma instrument which measured the distributions of electrons and ions over an energy range of 15 eV-30 keV and 20 eV charge-‘+0 keV charge- ‘. respectively (Paschmann er al., 1985). The plasma instrument was designed for optimum per- formance in magnetosphere and magnetosheath plasma. and therefore accurate ion densities and tem- peratures could not be obtained in the cold solar wind due to the relatively large spacing of the energy chan- nels ; however, good electron density (N) and ion bulk velocity (r,) measurements were obtained, which enabled the solar wind ram pressure (Nnz,ri) to be determined under the assumption that the solar wind consists almost entirely of protons.

In addition to the IRM data, we have used energetic

ion and electron particle flux measurements from two geosynchronous satellites (denoted 1977-007 and 1982-019) ; these satellites are located at U.T. +4.5 h and U.T. + 15 h, respectively. On each satellite, the

electron detector has six nested energy bins covering the range from 30 to 300 keV. The proton detectors on the two satellites differ slightly but in both cases contain five nested energy bins covering roughly the

range from 100 to 600 keV. (Exact energy ranges are given below in the appropriate figure captions. For

more information on these satellites, refer to Baker et

al., 1979, 1988.) Data from two equatorial magnetic observatories

have also been examined in order to monitor the reac-

tion of the dayside magnetosphere. The two ground stations, Hyderabad (148”E, 7.6”N geomagnetic;

78”E, 17”N, geographic) and Davao (195”E, 3.8”N geomagnetic ; 125”E, 7.O’N geographic) were selected such that one of them was at local noon during each of the two specific events studied.

Figure 1 shows the position of the two ground

stations and the three satellites at 04:OO U.T. and 07:OO U.T. on 5 September 1984, which are the approximate

times of two interplanetary discontinuities to be pre- sented below. The nominal positions of the mag-

netopause and bow shock are shown in Fig. 1.

OBSERVATIONS

The magnetic storm of 4-5 September 1984 has been intensely studied as part of the AMPTE mission and in fact was the basis for the first storm-time

measurements of the charge state and composition of the bulk of the ring current (Gloeckler et al., 1985).

Here we concentrate on a part of that storm, OO:O& l2:OO U.T. on 5 September. At this time the bow shock and magnetopause had already reached their most compressed positions and were gradually moving out- ward. However. the D,, index. reflecting enhanced ring current particle fluxes, was still decreasing during the first half of this period and did not reach its minimum value of - 120 nT until 07:OO U.T. on 5 September.

Figure 2 shows ground-based and IRM data for the period 00:0&12:00 U.T. on 5 September. The top seven pan& all represent IRM data: the magnetic field magnitude (2a), azimuth (2b). and elevation (2~) in GSE coordinates ; the angle between the magnetic held direction and the Earth-satellite line (2d), which should be approximately the unperturbed bow shock normal for this local time (near noon); and three plasma moments : electron density (2e) ; bulk ion vel- ocity (20 ; and the ion ram pressure (2g). Bow shock traversals can be most conveniently identified in the

Interaction of impulsive solar wind discontinuities with magnetosphere 843

-

2om1 0

20

.\ \

_ 1982-019 x f : IBM

x : ;

1977-007,” ,I I

/’ / /I , /

e*- .’ ,’

/, I I,

- 20

20-i

2 : 0

- i

T 1977-007 : x

1982-019 ,f’ /I

;

,/’ ,I’

FK. 1. THE POSITIONS IN THE GSE x--Y PLANE OF THE TWO

GROUND STATIONS AND FOUR SATELLITES AT TWO DIFFEREST

TIMES WHEN THE MAGNETOSPHERE WAS OBSERVED TO REACT

TO THE PASSAGE OF SOLAR WIND DISCONTlNUITlFS: (a) AT

03:35 U.T. ; AND (b) AT 06:46 U.T. ON 5 SEPTEMBER 1984.

bulk velocity trace, since the solar wind speed during this 12-h interval is 45@-500 km s ‘. while the velocity

of the shocked magnetosheath plasma is about 150 km s- ‘.

At the start of the interval (OO:OWO:27 U.T.). the spacecraft was in the magnetosheath at ,I’,;,, = II.3 R,, Yc;sE = 0.3 R,, Z,,, = 0.7 R,. The bow shock moved past the spacecraft three times between 0027 and 0051 U.T., after which the satellite remained in the solar wind for some time. (During this period, the bow shock was still recovering from an extremely compressed position during the unusual solar wind conditions of the previous day. when it was observed inside 10 R,.) From 0l:OO U.T. until 08:30 U.T., the ram pressure observed in the solar wind (Fig. 2g) remained relatively constant at I nPa. plus or minus 20%. Of course, there were some brief interruptions in the solar wind observations during this period. when the satellite passes into the magnetosheath. Also, a gradual depression in the kinetic pressure to

about 0.85 nPa occurred near 03:OO U.T., and there

were brief changes in the ram pressure at 03:35 U.T.

and 06:46 U.T. accompanying interplanetary dis- continuities, as discussed below. After 08:30 U.T., the

ram pressure increased again. This increase in pressure caused the bow shock to move radially inward away

from the satellite. and this effect. combined with the fact that the satellite was moving outward and leaving the normal range of bow shock distances for this local

time, explains why no further bow shock traversals were seen after 08:30 U.T.

During the period of relatively constant solar wind pressure (01:00-08:30 U.T.), IO bow shock traversals

were recorded. These are easiest to recognize in the

ion velocity (Fig. 2f), although they are also visible in the electron density (Fig. 2e). the ram pressure (Fig. 2g), and the magnetic field magnitude (Fig. 2a). They occurred in two distinct groups, there were four crossings during 03:45-04:45 U.T. and six cross-

ings during 06:5@08: I5 U.T. Apparently the satellite lay outside the bow shock during most of the 01 :O& O&30 interval, except during the two time intervals mentioned above, when some perturbation shifted the

bow shock outward to radial distances greater than the satellite for about I h each time.

Turning to the magnetic field data, the inter-

planetary magnetic field amplitude remained remark- ably constant with a value of 20 nT during the entire time period of the measurements, except for a gradual decrease in the field strength which set in at the very end (after I I :20 U.T.). As expected from the Rankine-

Hugoniot jump conditions for fast shock waves, large magnetic field increases characterized the transition from solar wind to magnetosheath. During 00:5& 03:33 U.T.. magnetic field fluctuations occurred with 6B of the order of a few nanoteslas. This signature is

typical of the region of upstream reflected ions known as the ion foreshock (e.g. Fairfield, 1969; Hoppe cr al.. 1981 ; and references therein). That the satellite was in the foreshock at this time is also indicated by the field direction, which was nearly parallel to the shock normal (Fig. 2d) and by energetic particle sig- natures shown by Mobius er al. (1986).

Figures 2b and c show that the IMF direction as well as magnitude stayed relatively constant for long periods, interrupted by occasional discontinuities. For

example, a discontinuity at 00:54 U.T. increased the IMF elevation angle from -45” to nearly 0 and coincided with a solar wind kinetic pressure decrease. The period of relatively constant solar wind ram pres- sure (Ol:O(M8:30 U.T.) was marked by two dis- continuities separated by nearly constant IMF. From OO:54 to 03:35 U.T.. the azimuth and elevation were both near zero degrees, implying that the IMF was

844 J. LABELLE ef al.

AMPTE/IRM 5 SEP 1984

AB

%3n

Ne* I Cdl

V P

[km/s1

UT

60

40

20

0:

90 :

O-

-90 y

45

0)

-45.

135:

90 ]

45

0

20

10

0

600

400

200

0 2.4

1.6

0.8

0: I I

I I 100 “T ,

t HYDERABAD H

0:oo 2:oo 4:oo 6:00 8:00 lo:oo 12:oo

‘GZ 12:07 rtx=

0.98 ny=

0.19 nz=

0.00 ‘K 13:20

FIG. 2. PLASMA AND MAGNETIC FIELV DATA FROM THE AMPTEjIRM SATELLITE DURING A PORTION OF AN

OUTBOUND ORBIT NEAR THE EARTH'S BOW SHOCK: (a) MAGNETIC FIELD MAGNITUDE: (b) MAGNETIC FIELD

AZIMUTH. IN GSE COORDINATES; (C) MAGNETIC FIELD ELEVATION, ALSO IN GSE COORDINATES; (d) ANGLE BETWEEN THE MAGNETIC FIELD AND THE EARTH-SATELLI~ AXIS, WHICH IS APPROXIMATELY IDENTICAL TO

THEANGLEBETWEENTHEFIELDANDTHEBOWSHOCKNORMALFORTHESEL~CALTIMES;(~)ELECTR~NDEN~ITY;

(f) ION BULKVELOCITY;AND(g)RAMPRESSURE(%I,~2). The bottom two panels display the H-components from two ground-based magnetometers : (h) Hyderabad (148 E. 7.6’ N geomagnetic) ; and (i) Davao (195 E. 3.8” N geomagnetic). Thirteen bow shock traversals are denoted by vertical lines, and three discontinuities in the magnetic field direction are indicated by dashed vertical lines. The two discontinuities accompanied by pressure pulses at 03:35 U.T. and 06:46 U.T.

each coincided with signatures in the geomagnetic data.

Interaction of impulsive solar wind discontinuities with magnetosphere 845

parallel to the typical bow shock normal for this local time (Fig. 2d). (During this time the azimuth rotated

gradually from about - IO’ to about +25”.) As men-

tioned above, the satellite was in the foreshock during this interval. At 03:35 U.T., the IMF elevation angle switched suddenly from near zero to about -75’, corresponding to a change of IMF direction from nearly parallel to nearly perpendicular to the unper- turbed bow shock normal. As expected, after this change the satellite was no longer in the foreshock

region, as evidenced by the absence of upstream 6B fluctuations. Subsequently, the IMF direction remained constant until a pair of closely-spaced dis-

continuities arrived at 06:42-06:46 U.T. The change in azimuth during 04:306:40 U.T. was not significant

since the azimuth angle has little significance when the elevation is near 90’. At 06:46 U.T. the IMF elevation switched from -90’ to 0 with an azimuth of -90’. corresponding to a rotation of the IMF in the plane tangential to the bow shock, i.e. a transition from one perpendicular bow shock geometry to another. The

field again remained approximately constant in direc- tion until after 09:OO U.T. In the last 3 h shown in Fig. 2, 09:0&12:00, several complicated changes in

the IMF direction occurred, but as the pressure was increasing, and the satellite was most likely well away from the bow shock at this time, these events are not

considered further. Significantly, the two discontinuities (at 03:35 and

06:46 U.T.) were coupled with: (1) a sharp increase of approximately a factor of two in the solar wind kinetic pressure lasting about l-2 min; (2) a slight decrease in the steady-state level of the solar wind

kinetic pressure (by about 20% in the 03:35 U.T. case and by only 8% in the 06:46 U.T. case); and (3) a cluster of bow shock crossings which appeared at IRM with a time delay of 8-12 min and lasted in each case

for about an hour. Since the satellite velocity was about 1 km SC’ outward, this was consistent with an

outward shift of the bow shock by at least 0.5-l .O RF. in about 8-12 min after each discontinuity, as might be expected qualitively from the decrease in the steady-state value of the solar wind dynamic pres- sure. The existence of multiple shock crossings in-

dicates, however, that the situation was somewhat complex, with inward/outward motions of the bow shock perhaps due to large scale undulatory waves on the shock surface or due to inward-outward motions of the shock as a whole.

The bottom two panels in Fig. 2 show the two ground magnetograms. Figure 2h is the H-component from Hyderabad which was near noon for the second discontinuity (06:46 U.T.), and Fig. 2i is the H-com- ponent from Davao which was near noon for the

first discontinuity (03:35 U.T.). The magneto-

grams showed similar features, although the Davao

data were of somewhat higher quality. Both solar

wind discontinuities and dynamic pressure pulses

(at 03:35 and 06:46 U.T.) were accompanied by a sudden impulse signature at each ground station with amplitudes of - IO nT in the first event and - 5 nT in the second event. The onset of these sudden impulse signatures occurred within 2 min of the observation

of the discontinuities/pressure pulses at IRM. The sudden impulse signature of the 06:46 U.T. dis-

continuity was distinctly weaker in both mag- netograms than that of the first discontinuity, an effect especially evident at Hyderabad, where the second

discontinuity was barely observable in the ground magnetic data. We have examined many other low-

latitude magnetograms from the dayside (e.g. Guam, Port Moresby, Memambetsu, Kanoya, Kakioka) ; all showed evidence of a sudden impulse at 03:35 U.T., and most though not all showed the 06:46 sudden impulse signature as well. After 03:35 U.T., the H- component in these equatorial magnetograms decreased, developing into a magnetic bay which cor- responded to an enhancement in the ring current seen

at geosynchronous orbit, as discussed below. The decrease was interrupted by an impulsive spike near 04:07 U.T. which was much more clearly expressed in

the Davao data than at Hyderabad ; this was coinci-

dent with the momentary inward motion of the bow shock detected l-2 min later by the IRM. There was also a short term increase of the H-component at both ground stations near 04:30 U.T., coincident with an inward motion of the bow shock recorded by the IRM. After that, the H-component at both locations continued to decrease, reaching a minimum value shortly after the second (06:46 U.T.) discontinuity;

thereafter, the H-component at both locations re- covered gradually, and no distinct features were seen in the magnetometer data to correspond to the

various changes in the bow shock position. Figure 3a.b (3c,d) shows the energetic ion and elec-

tron data from the daytime (night-time) geo- synchronous satellite. whose position during the two discontinuity/pressure pulse events is given in Fig. 1. As is particularly evident in the ion data, the energetic particle fluxes react similarly to both discontinuities. In the case of the first discontinuity, the picture is somewhat complicated because a dispersionless enhancement in both ions and the electrons occurred on the nightside first at 03:27 U.T., 8.5 min before the discontinuity. The slight increase in the dayside ions and electrons at or slightly before the discontinuity probably represents the dispersed signature of this injection as the particles drift around to the dayside.

846 J. LABELLE et al.

1977 -007 0500 9900 1300 LT

138

10:

105

IO5

loL IO'

7 105 %

7 103 m

n

‘= 0 10'

2 i G 108

g IO’

Lf z 106 -

105

10'

10'

105

IO"

IC’

~ b)

0200 0600 1000 UT

1982 - 019

0200 0600 1000 UT

FIG. 3. ENERGETIC PARTICLE DATA FROMTHETWOGEOSYNCHRONOUSSATELLITESFOR THEINTERVAL~~:~~-

12:OO U.T.ON 5 SEPTEMBER 1984: (a) INTEGRATED ELECTRON FLUXES IN SIX NESTED ENERGY BINS (30300, 45-300,6>300,95_300,I40300,AND 2003OOkeV) MEASUPEDFROMTHESATELLITE 1977~007.WHICH WAS

PRIMARILY ON THE DAYSIDE DURING THIS TIME; (b) INTEGRATED ION FLUXES IN FIVE NESTED ENERGY BINS

(145-560, 175-560, 215-560, 290560, AND 400560 keV) MEASURED FROM THE SAME SATELLITE;

(c) INTEGRATED ELECTRON FLUXES IN THE SAME ENERGY BINS AS (a), BUT FROM THE SATELLITE 1982-019, WHICH WAS ON THE NIGHTSIDE DURING THIS TIME; (d) INTEGRATED ION FLUXES FROM 1982-019, IN FIVE NESTED ENERGY BINS SLIGHTLY DIFFERENT FROM THOSE IN (b) (92600, 1288600, 195600, 254600, AND

377-600 keV). Vertical lines mark the times of the two interplanetary discontinuities detected by the IRM at 03:35 U.T.

and 06:46 U.T.

Aside from this complication, a repeatable signature occurred at both satellite positions : first, the ion and electron fluxes were depressed for l&30 min, and this was followed by an injection event in which the fluxes were enhanced. The injection event appeared first in the nightside data, without energy dispersion, and then appeared later on the dayside, with the higher energies enhanced first. The delay time in the case of the highest energies (300 keV electrons and - 600 keV ions) was very short, the order of a minute or so. The injection event coming after the first discon- tinuity was significantly larger than that following the second discontinuity. Not surprisingly, this global

ring current enhancement coincided with the magnetic

bay observed in the ground magnetograms.

DlSCUSSlON

It is interesting that the magnetospheric particle and geomagnetic signatures of the 03:35 U.T. dis- continuity are consistently larger than those of the later 06:46 U.T. discontinuity, despite the similarities of the two pressure pulses observed with IRM. One important difference, however, is that the first pressure pulse corresponded to a change from a radial to a perpendicular IMF orientation. while the second pres-

Interaction of impulsive solar wind discontinuities with magnetosphere 847

sure pulse did not. Fairfield et al. (1990) have sug-

gested that during radial IMF conditions the fore-

shock structure might effectively partially deflect the

solar wind from the front side of the magnetopause, while a perpendicular IMF orientation would not

cause this effect to the same degree. As a result, a shift from a radial to a perpendicular orientation would

expose the magnetopause to greater increased pres- sure than a similar shift between two perpendicular orientations. This could explain why the first dis- continuity was apparently so much more effective at

generating magnetospheric effects. Another factor important for the geosynchronous particle effects is

the state of the global ring current. The second event occurred during the height of the disturbance, and the already disturbed ring current at this time may explain why the particle signature of the discontinuity/pres- sure pulse was less distinct. Another factor impor- tant for the geomagnetic effects is the longitude at the dayside magnetic observatories at the time of the pressure pulses, i.e. the U.T. of the pressure pulses. This effect arises due to the importance of the equa- torial electrojet in amplifying low-latitude geo- magnetic signatures of magnetospheric compression (e.g. Sugiura, 1953). Since the electrojet current has a longitudinal dependence (stronger in Peru ; weaker in

the Asian sector) (e.g. Rastogi, 1962), this produces a U.T.-dependence in the efficiency of the generation

of geomagnetic signatures by solar wind pressure pulses. In the present case, both pressure pulses occurred when the Asian sector was on the dayside. The latitude of the geomagnetic observatories is of course also a factor, since the electrojet is latitudinally limited to a few hundred kilometers. This may explain why the Hyderabad station (latitude of 7.6”) detected weaker geomagnetic signatures than the Davao station (latitude of 3.8”).

There is a great deal of literature linking sudden impulses in low-latitude magnetograms to the passage of interplanetary discontinuities (reviews by Akasofu et al., 1977 ; Smith et al., 1986; and references therein). Based on many cases, an empirical relation has been developed relating the solar wind pressure (Pkln) to the local-time averaged magnitude of the sudden impulses at low latitudes (Is,) (Siscoe et al.. 1968;

Burlaga and Ogilvie, 1969; Nishida, 1978 : Smith et al.. 1986) :

I,, = (3.6 x 10’) (P:;,$ - P;,,$), (1)

where Pkln is expressed in Pascals and Is, is expressed in gammas. The local time-averaged Is, should be appropriate for comparison with our data since we observe about the same low-latitude magnetic field signature at many local times separated by hours.

Taking L., z 1.1 nPa and P,,,,, z 2.2 nPa (Fig. 2), we obtain approx. 12 nT. This is somewhat larger than

the observed enhancement, especially in the second

event (06:46 U.T.). Closer agreement is not really to

be expected since the geomagnetic response is affected by a number of factors such as IMF direction and the latitude and longitude of the particular ground station, as discussed above. Also, the short-lived

nature of the observed pressure pulses (_ 1 min) may imply that they are less effectively coupled to the

sudden impulses than were the mostly step-function cases which led to the empirical relation (1).

It is instructive to consider the time delays between

some of the observed events. There is approximately

a 2-min delay between the observed pressure pulses at IRM and the sudden impulse in the equatorial field.

This is reasonably well explained assuming that the pressure pulse travels with the solar wind or mag- netosheath velocity up to the magnetopause and then with the fast mode speed in the magnetosphere. Tak- ing a fast mode speed of about 1000 km s- ’ as in the cases studied by Wilken et al. (1982), we obtain a time delay of about 110 s for the geometry of Fig. 1. This is of course only a rough guess, since other factors play a role such as the orientation of the IMF dis- continuity/pressure pulse and where exactly it strikes

the magnetopause. The decrease in the flux of the particles at geo-

synchronous orbit sets in within a few minutes after each discontinuity and lasts 10-30 min. This may be

interpreted as the signature of the “growth” phase (McPherron, 1970) or “cigar” phase (Baker et al., 1977, 1978) of a substorm, in which the tailward

stretching of the magnetic field before the onset of the injection event leads to a change in the anisotropy

and to a decrease in the observed fluxes of particles

at geosynchronous orbit (see, for example, Fig. 2a of Baker, 1984). Such a decrease occurs first on the night- side, and should not be confused with a decrease

resulting purely from an adiabatic compression of the magnetosphere by the pressure pulse. The latter signature may not be observable in this case, probably because it is overwhelmed by substorm effects. [As pointed out by Akasofu (1977) to observe such purely compressional effects, the substorm effects must be absent. This is certainly not the case for 5 September 1984, for which substorm effects are dominant.] The comparison between the dayside and nightside data is also complicated by the fact that drifting particles originating at geosynchronous orbit on the nightside reach orbits either inside or outside geosynchronous orbit on the dayside, depending on the pitch angle (Paulikas and Blake, 1979), but the data seem con- sistent with a disturbance that occurs first on the

848 J. LABELLE et al.

nightside and then drifts to the dayside. This hypoth- very disturbed, and injection events occur which are

esis, that the particle signatures observed at geo- not associated with the discontinuities, such as that at

synchronous orbit indicate the growth phase of a sub- 03:27 U.T. However, the fact that substorm growth

storm, is supported by previously published evidence phase onsets immediately follow each discontinuity,

that IMF changes trigger substorms (e.g. Rostoker, and that particle injection events follow with a reason-

1983). able time delay, is strong evidence for a connection.

An alternative explanation for the flux decreases seen in the daytime geosynchronous particle fluxes

invokes the loss of energetic ions with large pitch

angles when they drift into the magnetopause. It is

clear from the ground magnetograms that some sort of compression of the magnetosphere occurs in

response to the pressure pulses/IMF discontinuities. This effect must move the magnetopause inward. As a result, one expects to observe “shadowing” of large pitch-angle energetic ions (electrons) in the morning

sector (afternoon sector), since energetic ions drift from dusk to dawn across the dayside, while energetic electrons drift from dawn to dusk. The result is a complex set of signatures in the fluxes of particles as a function of pitch-angle and energy (Wilken et al.. 1982, 1986), which has been used effectively as

a remote sensing technique for detecting the mag- netopause and its radial motions (Baker et al., 1988). This hypothesis would explain why the daytime flux

decreases are much clearer in the ions than in the electrons, since the daytime geosynchronous satellite (1977-007) was on the morningside, putting it in the “shadowed” region for ions but not for electrons. Hence, while the nightside geosynchronous particle

signatures seem most likely to result from growth phase substorm effects, the dayside particle signatures could result from the same effects, or from the “shadowing” of the energetic ions on the morningside

outer magnetosphere due to the tendency of the large pitch-angle energetic particles to be lost at the mag- netopause. or from a combination of both effects.

In both events studied, the bow shock is observed to move from inside the IRM position to outside for

periods of about I h (03:35 U.T. event) and I.5 h (06:46 U.T. event). Since the satellite outward velocity

is about I km SC’, this is consistent with outward

shifts in the bow shock position of at least 36OG5400

km. In the first event, a fairly large (- 20%) reduction in the solar wind pressure accompanies the dis-

continuity. and this would imply a 3% outward per-

turbation in the bow shock’s position-about 0.5 RE

or 3300 km, in good agreement with the above esti- mate. However. the small reduction in solar wind pressure across the second discontinuity, about 8%, would be expected to produce only a 1500 km change in the shock’s position. This seems inconsistent with

the observed perturbation. Of course. there could be more severe changes in the solar wind pressure during the time that the IRA4 is in the magnetosheath ; these

cannot be observed but could cause the bow shock to move further outward. Another possibility is that the

bow shock reacts not just purely to pressure changes but also to changes in the IMF direction independent

of the pressure. as has been suggested in a number of papers (Auer, 1974; Volk and Auer, 1974; Neubauer, 1975). A third possibility for the large excursion in the bow shock position is that the pressure pulse sets

the shock into an oscillatory motion. Examples of such oscillatory motions of the bow shock have

been reported (e.g. Kaufmann. 1967 ; Freeman ef al.. 1967).

The time delay between the discontinuity and the particle injection falls in the range of time delays between substorm onsets (defined by the onset of Pi2 pulsations) and sudden impulses (Iyemori and Tsunomura, 1983; Baumjohann, 1986). The time delay also compares favorably with those determined between sudden commencements and triggered mag- netospheric particle bursts (Tholen et al., 1986). Other works have also provided evidence for the linkage between sudden impulses in the geomagnetic field. their associated interplanetary discontinuities. and geomagnetic storms or substorms (e.g. Schieldge and Siscoe. 1970; Burch. 1972 ; Ijima, I973 ; Kokubun et al.. 1977; review by Akasofu, 1977; and references therein). However, for the same reason, we cannot be absolutely certain that the discontinuity/pressure pulse events trigger substorms, since the situation is

In summary. we have examined data from satellites and magnetic observatories during an unusual period for which the solar wind pressure and magnetic field are relatively constant except for two discontinuities with associated - I min duration pressure pulses. The observations indicate: (I) the bow shock reacts rapidly to solar wind dynamic pressure variations; (2) both discontinuity/pressure pulse events were fol- lowed more or less immediately by a decrease in the geosynchronous particle fluxes indicating the growth phase of a substorm and possibly also the shadowing of energetic ions by the compressed magnetopause, followed after IO-30 min by a sudden enhancement in the energetic particles which began on the nightside and drifted to the dayside; and (3) sudden impulse ground signatures are observed which seem to depend critically on the IMF orientation, as well as on the longitude and latitude of the particular ground

Interaction of impulsive solar wind discontinuities with magnetosphere 849

stations. The magnetospheric signatures of the first discontinuity were significantly larger than those of

the second one, despite the fact that the associated solar wind pressure pulses were approximately equal.

This could be explained by the IMF direction, by the

difference in the disturbance level of the ring current at the times of the two events, or by longitude effects which would affect the ground signatures.

Acknowle&emenrs-The entire AMPTEIIRM plasma and magnetic field teams are to be thanked for their contributions to the measurements. We wish to thank G. Paschmann, N. Sckopke, and I. Papamastorakis for helpful discussions, and we are grateful to H. Liihr for use of the magnetometer data. We also thank R. D. Belian of Los Alamos National Laboratory (LANL) for data used in this study, and we thank C. Baca and K. Sofalv of LANL for data analvsis support. The work at Dartmouth College was supported by NASA contract NAGW-1540. The work at the Applied Physics Laboratory was supported by NASA under Space and Naval Warfare &stems Command contract N00039-87- C-5301 of the Navy. .

REFERENCES

Akasofu, S.-I. (1977) Physics of Magnetospheric Substorms, p. 190. Reidel, Dordrecht, Holland.

Auer, R. D. (1974) Magnetohydrodynamic aspects of the Earth’s bow shock I. Equilibrium bow shock position. J. geophys. Res. 79. 5 118.

Baker. D. N. (1984) Particle and field signatures of substorms in the near magnetotail. in Magnetic Reconnecrion in Space and Laboratory Plasmas, Geoph,vs. Monogr. 30 (Edited by Hones. E. W.). p. 193. AGU, Washington. D.C.

Baker. D. N., Belian, R. D., Fritz, T. A.. Higbie. P. R.. Krimigis, S. M., Sibeck. D. G. and Zwickl. R. D. (1988) Simultaneous energetic particle observations at geo- stationary orbit and in the upstream solar wind : evidence for leakage during the magnetospheric compression event of November I. 1984. J. &ophy.~. Res. 93. 14317.

Baker. D. N.. Belian. R. D.. Hiebie. P. R. and Hones. E. W. (1979) High-energy magnetospheric protons and their dependence on geomagnetic and interplanetary con- ditions. J. geophvs. Res. 84. 7138.

Baker, D. N., Higdie, P. R.. Hones. E. W. and Belian. R. D. (1978) High-resolution energetic particle measurements at 6.6 R, 3. Low energy electron anisotropies and short term substorm predictions. J. yroplt~~s. Res. 83, 4863.

Baumjohann, W. (1986) Some recent progress in substorm studies. J. Geomug. Geoelecr. 38. 633.

Burch. J. L. (1972) Preconditions for the triggering of polar magnetic substorms by storm sudden commencements. J. geophys. Res. 77. 5629.

Neubauer. F. M. (1975) Nonlinear oblique interaction of interplanetary tangential discontinuities with magneto- gas-dynamic shocks. J. geophvs. Res. 80. 1213.

Nishida. A. (1978) Geomagnetic Diagnosis of the Ma,q- nerosphere, p. 2 I. Springer. New York.

Paschmann. G.. Haerendel. G.. Sckopke. N., Mobius. E., Liihr. H. and Carlson. C. W. (1988) Three-dimensional plasma structures with anomalous flow directions near the Earth’s bow shock. J. geoph,vs. Res. 93. 11279.

Paschmann. G., Loidl. H.. Gbermayer, P.. Ertl. M.. Labor- enz. R.. Sckopke, N., Baumjohann, W.. Carlson. C. W. and Curtis. D. W. (1985) The plasma instrument for the AMPTE IRM. IEEE Truns. Geosci. Remole Sensitq GE- 23. 262.

Burgess, D. and Schwartz. S. J. (1988) Colliding plasma structures: current sheet and perpendicular shock. J. geopl1y.v. Res. 93. 1 1327.

Burlaga, L. F. and Ogilvie. K. W. ( 1969) Causes of sudden commencements and sudden impulses. J. yeoph~~s. Res. 74. 28 15.

Paulikas. G. A. and Blake, J. B. (1979) Effects of the solar wind on magnetospheric dynamics : energetic electrons at synchronous orbit. in Qtran/i/uriw Modeling o/ Ma,y- m~rospheric Processes (Edited by Olson. W. P.). AGU. Washmgton. D.C.

Fairfield, D. H. (1969) Bow-shock associated waves observed in the upstream interplanetary medium. J. ycwph,~.s. Rrs. 74, 3541.

Potemra. T. A.. Liihr. H.. Zanetti, L. J.. Takahashi. K., Erlandson, R. E.. Marklund. G. T.. Block. L. P.. Bloom- berg. L. G. and Lepping. R. P. (1989) Multi-satellite and ground-based observations of transient ULF waves. J. ,yeophy.r. Res. 94. 2543.

Fairfield. D., Baumjohann. W.. Paschmann. G.. Liihr. H. Rastogi. R. G. (1962) Longitudinal variation in the equa- and Sibeck, D. (1990) Upstream pressure variations associ- torial electrojet. J. ammos. lerr. Phys. 24. 1031.

ated with the bow shock and their effects on the mag- netosphere, J. qeophvs. Res. 95. 3773.

Freeman. J. W.. Kavanagh, L. D. and Cahill, L. J. (1967) An observation of transient variations in the magnetospheric boundary position. J. geophvs. Res. 72, 2040.

Friis-Christensen, E., McHenry, M. A., Clauer, C. R. and Vennerstrom. S. (1988) Ionospheric traveling convection vortices observed near the polar cleft : a triggered response to sudden changes in the solar wind. Geophys. Res. Left. 15, 253.

Gloeckler. G., Wilken. B., Stiidemann. W.. Ipavich, F. M., Hovestadt, D.. Hamilton, D. C. and Kremser. G. (1985) First composition measurement of the bulk of the storm- time ring current (i-300 keV/e) with AMPTE/CCE. Geophys. Res. Left. 12, 325.

Hoppe. M. M.. Russell, C. T.. Frank. L. A., Eastman, T. E. and Greenstadt, E. W. (1981) Upstream hydromagnetic waves and their association with backstreaming ion popu- lations : ISEE-I and -2 observations. J. geophys. Res. 86. 4471.

Ijima, T. (1973) Interplanetary and ground magnetic con- ditions preceding SSC-triggered substorms. Rep. Ionosph. Space Res. Japan 27. 205.

Iyemori. T. and Tsunomura. S. (1983) Characteristics of the association between an SC and a substorm onset. Mem. natn. Inst. Polar Res.. Spec. Issue 26, 139.

Kaufmann, R. L. (1967) Shock observations with the Explorer 12 magnetometer. J. geoph,vs. Res. 72. 2323.

Kokubun, S., McPherron, R. L. and Russell, C. T. (1977) Triggering of substorms by solar wind discontinuities. J. ,yeoph.vs. Res. 82. 74.

Liihr. H., Klocker, N., Oelschllgel, W.. HHusler, B. and Acuna. M. (1985) The IRM flux-gate magnetometer. IEEE Trans. Geosci. Remote Sensing GE-23, 259.

McPherron, R. L. (1970) Growth phase of magnetospheric substorms. J. geophys. Res. 75. 5592.

Mobius. E., Hovestadt, D., Klecker: B.. Scholer. M.. Ipavich, F. M.. Carlson. C. W. and Lin. R. P. (1986) A burst of energetic O+ ions during an upstream particle event. Geoppl~~~. Res. Lerr. 13. 1372.

850 J. LABEI

Rostoker, G. (1983) Triggering of expansive phase inten- sifications of magnetospheric substorms by northward turnings of the interplanetary magnetic field. J. geophys. Res. 88, 698 1.

Schieldge, J. P. and Siscoe, G. L. (1970) A correlation of the occurrence of simultaneous sudden magnetospheric com- pressions and geomagnetic bay onsets with selected geo- physical indices. J. atmos. (err. Phvs. 32, 1819.

Schwartz, S. J., Chaloner, C. P., Hall, D. S., Johnstone, A. D.. Rijnbeek, R. P., Southwood, D. J. and Woolliscroft, L. J. C. (1985) An active current sheet in the solar wind. Nature 318, 269.

Schwartz, S. J., Kessel, R. L., Brown. C. C., Woolliscroft, L. J. C.. Duntoo. M. W.. Farrueia. C. J. and Hall. D. S. (1988) Active current sheets nea; the Earth’s bow shock. J. geophys. Res. 93, 11295.

Sibeck, D. G.. Baumjohann, W., Elphic. R. C., Fairfield, D. H., Fennell, J. F., Gail, W. B., Lanzerotti, L. J., Lopez, R. E.. Luehr. H.. Lui. A. T. Y.. Maclennan, C. G.. McEntire. R. W., Potemra. T. A., Rosenberg, T. J. and Takahashi, K. (1989a) The magnetospheric response to 8- minute period strong amplitude upstream pressure vari- ations. J. ,qeophys. Res. 94. 2505.

Siscoe, G. L., Formisano, V. and Lazarus, A. J. (1968) Relation between geomagnetic sudden impulses and solar wind pressure changes-an experimental investigation. J. geophys. Res. 13. 4869.

Smith. E. J., Slavin, J. A., Zwickl. R. D. and Bame, S. J. (1986) Shocks and storm sudden commencements. in Solar

_LE et al.

Wind-Magnetosphere Coupling (Edited by Kamide, Y and Slavin, J. A.), pp. 34S365. Terra Scientific, Tokyo.

Suguira, M. (1985) The solar-diurnal variation in the ampli- tude of sudden commencements of magnetic storms at the geomagnetic equator. J. geophys. Res. 58, 571.

Tholen, S. M., Armstrong, T. P. and Lanzerotti, L. J. (1986) Correlation of geomagnetic storm sudden commencement with triggered magnetospheric particle bursts, J. geophys. Res. 91, 13717.

Thomsen. M. F., Gosling, J. T., Bame, S. J., Quest, K. B.. Russell, C. T. and Fuselier, S. A. (1988) On the origin of hot diamagnetic cavities near the Earth’s bow shock. J. geophvs. Res. 93. 113 1 1.

Thomsen, M. F., Gosling, J. T., Fuselier, S. A., Bame, S. J. and Russell, C. T. (1986) Hot diamagnetic cavities upstream from the Earth’s bow shock. J. geophys. Res. 91, 2961.

Viilk, H. J. and Auer, R. D. (1974) Motions of the bow shock induced by interplanetary disturbances. J. geophys. Res. 79, 40.

Wilken, B., Baker, D. N., Higbie, P. R., Fritz, T. A., Olson. W. P. and Pfitzer, K. A. (1986) Magnetospheric con- figuration and energetic particle effects associated with an SSC : a case study of the CDA W 6 event on March 22, 1979. J. geophvs. Res. 91, 1459.

Wilken. B.. Goertz, C. K.. Baker, D. N., Higbie, P. R. and Fritz, T. A. (1982) The SSC on July 29, 1977 and its propagation within the magnetosphere. J. geophys. Res. 87, 5901.