Planet. Space Sci. 1975, Vol. 23, pp. 693 to 711.
Persamon Press. Printed in Northern Ireland
IN THE MAGNETOSPHERE
C. C. LJN and L. J. CAHILL, Jr. School of Physics and Space Science Centre, University of Minnesota, Minnesota 55455, U.S.A.
(Received injinal form 16 September 1974) Abstrac,Peveral substorms were observed at Explorer 45 in November and December 1971, and January and February 1972, while the satellite was in the evening quadrant near L = 5. These same substorms were identified in ground level magnetograms from aurora1 zone and low latitude stations. The satellite vector magnetic field records and rapid run ground magnetograms were examined for evidence of simultaneous occurrence of Pi2 magnetic pulsations. Pulsations which began abruptly were observed at the satellite during 7 of the 13 substorms studied and the pulsations occurred near the estimated time of substorm onset. These 7 pulsation events were also observed on the ground and 6 were identified in station comments as Pi2. All of the events observed were principally compressional waves, that is, pulsations in field magnitude. There were also transverse components elliptically polarized counterclockwise looking along the field line. Periods observed ranged from 40 to 200 SW with 80 set often the dominant period.
term magnetospheric substorm was introduced when it became apparent that aurora1 zone magnetic disturbances (polar magnetic substorms) and intense aurora1 displays (aurora1 substorms) were part of much larger scale events involing the night side of the outer magnetosphere and the magnetotail (Akasofu, 1964; Coroniti, et al., 1968). These events, with typical duration of an hour or two, occur, on the average, about once each day. They produce large irregular magnetic disturbances at aurora1 latitudes and have strong local time dependence. Often in a disturbed period there are a sequence of closely spaced substorms. In contrast, magnetic storms which seem to develop from a series of substorms, involve the inner magnetosphere and exhibit a uniform magnetic depression at low latitudes and all local times. They occur only once or twice a month and last for a day or two. Magnetospheric substorms seem to be the result of introducing solar wind energy inside the magnetosphere. This deposition of energy leads to a related group of dynamic processes of the magnetosphere-changes in plasma flow, inflation of the magnetic field, changes in energetic particle flux, precipitation of particles into the high latitude ionosphere, auroras, intense ionospheric currents and occasionally to the establishment of a symmetric ring current at L = 24. The abundance of phenomena, the diversity of form and the large scale of these dynamic changes have made the magnetospheric substorm a challengingproblem to observation and theory. Many groups are studying different aspects of substorms and there have been a few studies of simultaneous magnetospherewide phenomena (Whalen and McDiarmid, 1972; Choy et al., 1971; Rostoker, 1968; DeForest and McIlwain, 1971; Lezniak and Winckler, 1970; McPherron, 1973). One problem in tying together observations of a substorm at many separated locations is the timing. The expansion phase of the substorm is the time of most rapid and large change. The earlier growth phase and later recovery phase may occur over longer periods of time. During these phases there are no effects or small effects at some locations in the magnetosphere and even when these phases have observable effects the changes are gradual with indistinct beginnings and endings. Even the expansion phase appears to occur at different times depending on location. Saito (1961) was first to note that Pi oscillations accompany the onset of a substorm on ‘a one-to-one basis. He showed that whenever a Pi oscillation is observed at low-latitude stations, a substorm has just occurred somewhere In 693 0 The
C. C. LIN
and L. J. CAHILL,
the aurora1 zone. Rostoker (1968, 1972) has suggested that Pi2 pulsations occur almost simultaneously at separated ground magnetic observatories and may be useful as an indicator of the onset of the expansive phase. Sakurai (1970) and Smith (1973) have considered the source of this pulsation to be the outer magnetosphere near the midnight meridian. Pi2 pulsations are defined as irregular pulsations of period 40-150 set (Jacobs et al., 1964); some workers extend the upper limit to 300 set (Rostoker, 1967; Saito and Matsushita, 1968). Amplitudes observed on the ground at low latitudes range from 1 to 5 y. In the aurora1 zone, the Pi2 may have amplitudes as high as 100 y. They have been associated with magnetic disturbance at high latitudes (Jacobs, 1970; Campbell, 1967). Pc4 and Pc5 pulsations with periods, 45-150 set and 150-600 set respectively, are classif?ed as continuous pulsations of longer duration than Pi2 with amplitudes of l-10 y on the ground. Pc4 and Pc5 are observed more commonly in magnetically quiet periods. The distinction between PC and Pi pulsations is difficult to make in Explorer 45 data since we rarely observe long continuous trains of pulsations that might be identified as PC4 or Pc5. Most of the pulsations observed are small amplitude, irregular and last for only a few cycles. Long duration events have been seen, however, in the DODGE satellite records at geostationary orbit (Dwarkin et al., 1971). In this paper we describe observations of Pi2 pulsations in the magnetosphere. The observations were obtained with the Explorer 45 fluxgate magnetometer during late November and December of 1971, and January and early February of 1972, when the satellite was in the evening quadrant (Cahill, 1973). The observations we used were obtained near apogee between L = 4.8 and 5.5 (where the fluxgate magnetometer was in high gain and the sensitivity was better than 1 y) and within 15’ of the magnetic equator. Aurora1 magnetograms (principally College (CO), Fort Churchill (FC), Baker Lake (BL), Great Whale River (GWR)) and low-latitude magnetograms (Guam (GU), Honolulu (HO), San Juan (SJ)) were also examined for evidence of substorms. We searched the ground magnetograms for evidence of substorms during the intervals when the satellite was near apogee (three 3-hr intervals each day). Usually, but not always, there were corresponding large field distortions at the satellite during the substorms. Some indications of the expansion phase of the substorm, sudden and large field changes, were often obtained from the ground and satellite records. We then searched for evidence of mi~ropulsations in high time resolution plots of the satellite data and in rapid run ground magnetograms. Finally we examined the station comments that accompany the ground magnetograms. Sometimes pulsations in the ground magnetograms are identified as Pc4, Pc5, or Pi2 on the magnetogram copies obtained from World Data Center A. Thirteen clearly identified substorms occurred during the 3 month interval of this study while the satellite was in the right location (near apogee when the fiuxgate ma~etome~r was in high gain state) to provide observations (Table I). Nine of these substorms produced distinct field distortions at the satellite (A.B > 15 y, INC change > 4” or DEC change > lo; changes are listed in Table 1). One of the thirteen substorms, 3 January 1972, had no clear indication of pulsations at the satellite. Pulsations were observed at the satellite during 12 substorms. Seven of these 12 had distinct, near simultaneous, pulsation onsets at the satellite and in ground records and we assumed they were Pi2 pulsations. These are labeled Pi2 in Table 1. All of these seven, except 11 February 1972, were identified as Pi2 by at least one ground observatory. Five others without distinct onsets were considered less convincing as evidence of Pi2 in the magnetosphere because we were less certain that the ground observatories and the satellite were observing the same event.
11:30-13:OO 08:[email protected]
:00 21:00-23:30 03:30-05:30 lO:OO-14:OO lO:OO-12:Oo 16:30-18:30
22 Nov. 1971
1971 1971 1971 1971 1971 1971 1971
1971 1972 1972 1972 1972
23 Nov. 9 Dec. 12 Dec. 17 Dec. 18 Dec. 19 Dec. 19 Dec.
24 Dec. 3 Jan. 4 Jan. 15 Jan. 16 Feb.
05:30-80:00 O5:OQ-OQ:OO 06:00-08:00 12:30-14:OO 02:00-04:00
Time interval (UT)
121 152 155 190 288 :50 6
::: 25 15
1:; 106 107 1 4 1 10 2
8 5 9 15 12 12 4
20 40 35
26 75 86
1; 7 1
DEC ([email protected]
INC ([email protected]
Magnetic field distortion at Explorer 45
70-l 10 70-100 70-120 160 40-130
120-180 55-200 o-160 50-120 70-120 50-100 50-90
l-2 2-3 2
2-3 4 l-3 24 224 l-2 l-2
Peak to peak amplitude (y)
TABLE 1. SUB~TORMS CONSIDERED FOR EVIDENCE OF Pi2 PULSATION$
Pi2 Associated with SSC Activity begins at about 11: 50 UT Pi2 Very quiet before 22:00 UT Pi2 Activity begins at about 1O:OOUT Activity begins at about 1O:OOUT Occurs approximately between 17:20-18:40 UT Pi2 Noisy data, no pulsations observed Pi2 Pi2 Pi2
3 3 g
z G Remarks
22 75 99
155 190 288
(1971) 22 Nov. 06:35 9 Dec. IO:00 17De.c. 04:16
(1972) 4Jan. 06:08 15 Jan. 12:48 16 Feb. 02:35
Event Date UT
2055 2020 1850
2300 2147 2133
5.1 5.2 5.4
80, 180 55-200 50-120
2 6 2
Compressional Not available Anti-clockwise
Not available Anti-clockwise Anti-clockwise
Sitka Honolulu Fredericksburg Great Whale River
Boulder Sitka Boulder
110-130 50-80 60-80 120
90-100 110 50
65 2 5 IO-25
5 5 5
Sitka Honolulu None
Honolulu Honolulu Honolulu: A series of pulsations from 19:03 UT, 16 Dec. through OS:00 UT, 17 Dec. was identified as Pi2 Sitka
Stations identify Pi2
Pi2 PULSATIONS IN THE MAGNETOSPHERE
The seven Pi2 events, selected as described above, will be described in detail in the next section. A summary is provided in Table 2. In each case at least one near simultaneous distinct onset of Pi2 at the satellite and on the ground occurs near the sudden onset of the expansion phase of a substorm. There are other pulsations observed that may not be associated with a distinct substorm onset. In the last section we consider the significance of these events with regard to the use of Pi2 as an indicator of expansion phase onsets and with regard to the generation and propagation of Pi2. 2. RESULTS
The satellite records of micropulsations are expressed in an orthogonal coordinate system; hBo positive towards the north, AB.+positive towards the east and A& positive towards the Earth. Each AB component represents the measured field component minus the predicted component (Cain and Sweeney, 1970). The original satellite measurements were taken approximately once each second. An 8 second average was calculated. A 15 point (600 set) sliding average was subtracted from the data. Finally, a 3 point (24 set) sliding average was calculated for the satellite data points shown here. The magnetic field substorm distortion is shown by inclination (INC), declination (DEC) and field magnitude AB, where INC is positive when the radial component is towards the Earth, DEC is positive when the 4 component is towards the east, and AB is the difference between the magnitude of the observed field and the magnitude of the predicted field. Near the geomagnetic equator, the main field lies near the 8 direction, and AB changes can be interpreted as principally due to fluctuations in the 19component. Both AB and A& are shown, however, since AB is calculated from magnetometer measurements only. A& calculation requires computations involving satellite spin axis direction measurements which may introduce additional errors. 4 January 1972 (06 : 08 UT) A Pi2 wave event was observed on 4 January 1972. Plots of the magnetic field data for the substorm associated with this Pi2 are shown in Figs. I(a) and (b). The records from Fort Churchill, located near local midnight (and Great Whale River) revealed that the possible onset time of substorm expansion was approximately at 06:08 UT. The field magnitude AB observed on the satellite showed a similar change to that at Fort Churchill. Note that a small fluctuation in field magnitude appeared at about 06:08 UT. The micropulsation activity around the time interval of interest is shown in Fig. l(c). The records from Sitka, near 20:00 MLT indicate that the Pi2 began at 06:08 UT, lasted in two or more bursts for about 20 min and was predominantly in the Hand D components. During the event Sitka (L = 4) was about 0l:OO MLT to the west of the foot of the geomagnetic meridian through Explorer 45 (L = 5.2). The fluctuations observed on the satellite, mainly in the A& (or AB) component, started at the same time as recorded on the ground and had approximately the same time duration. The long period oscillation in AB, started well before 0608 UT and is not related to the Pi2 event. There was virtually no AB+ component. The Pi2 oscillations had a period of cu. 125 set and peak to peak amplitude of 6.5 y in H and D components (at Sitka) while the apparent period and amplitude seen on the satellite in the A& component were 80 set and 2 y respectively. Power spectra were prepared from the satellite data using Fast Fourier Transform techniques. The length of the time series for each individual Pi2 event was taken to be approximately 17 min of data.
C. C. LIN and L. J. CAHILL, JR
-co FC L MLAT
GU 200 L INC
I - DEC SJ
FIG. l(a). GROUND MAGNET~GRAMS KJR THB H OR x COMPONENT DURING A SUBSTORMTHATOCCURREDON 4 JANUARY 1972. Local midnight is indicated by an arrow. The vertical dashed line shows the onset of the substorm expansion determined from Pi2 activity.
o700 0600 2049
Fro. l(b). MAGNETIC RECORDS PROM Ewroaaa 45 (ORBIT 155) FOR THE SAME EVENT SHOWN IN FIG. 1(a). The horizontal scale of the figure is UT of observation; MLT, magnetic latitude and L value of the observationsarealsogiven. ABistheobserved magnetic field magnitude minus the predicted magnitude. The theoretical and experimental values of inclination and declination are shown by dashed and solid lines respectively.
The ramp on which the Pi2 rides (i.e. the bay) was first partially removed from the data by taking a 75 point sliding average before preparing the power spectra. The power spectra shown in Fig. l(d) demonstrated that for the ABB component there was a strong, broad peak from 70 to 120 set, with individual peaks centred at 80 and 100 sec. The intensities in the AB, and AB, components were insignificant compared with the A& component except the small peak at 43 set and one above 200 set (which may be spurious). At College (L = 5.4), about 0190 MLT to the west of Sitka, the magnetic records exhibited similar oscillations that started at about the same time with period about 120 seconds. Theamplitude of the H component reached 15 y at 06 : 12 UT. Pi2 activity was also observed on theground at Boulder, Fredericksburg and San Juan. Note that the Pi2 observed on the satellite had a very strong component along the main field (compressional wave).
IN THE MAGNETOSPHERE
699 PERIOD (WC)
Ah .-. \ i
JAN 4.1972 I5Y
Explorer 45 Orbit IS5 .Jlln 4.1972
'5Y 45 AB C~-..r‘c------./---_ 1 I4Y I A& I y--14Y
I I 1 I J.;-.-I I I
As, 14Y ABe I4Y I.1
FIG. l(c). MICROPU~~ATIONDATAFROM RAPID RUN MAGNETOGRAMSANDPROM ~LORRR45ASSOcIATEDWITHTHESIJRSTORM SHOWNIN FIGS. l(a) AND (b). The onset of Pi2 activity at approxi-
06 :08 UT is indicated vertical dashed line.
FIG. l(d). POWERSPECTRA PORTHE INTERVAL 06:0806:25 UT FOR THE EWLORER 45 MICROPULSATION DATA SHOWN IN FIG. l(c).
24 December 1971 (05 : 56,06 : 22 and 06: 37 UT)
On 24 December 1971 three Pi2 events separated by 26 and 17 min and each lasting for about 10 min, were observed both on the ground and at the satellite. The associated sub-storms are shown in Fig. 2(a). Onset of one substorm expansion could be inferred at --06 : 37 UT at College and also at Baker Lake which was near local midnight. The records from other observatories and Explorer 45 indicated no abrupt field change at the same time. It was possible to infer onsets from the satellite AB or from Fort Churchill at 05: 56, 06 : 22 UT and several other times. It was difficult in this multiple substorm event to be sure of the onsets of substorm expansions from normal run magnetograms or from the gross distortions of magnetic field at the satellite. Figure 2(b) shows the Pi2 micropulsations observed during the period of possible substorm onsets. Three Pi2 events occurred separately at 05: 56, 06:22 and 06:37 UT. There was little time delay (less than one minute) between the events recorded on the ground and observed at the satellite. It is quite common that Pi2 micropulsations observed on the ground appear in a series. The records at Sitka, about one hour magnetic local time to the west of the meridian through Explorer 45 (L = 5*2), showed that the oscillations were in the H and D components. The first two Pi2 bursts had periods of 110-120 set and peak to peak amplitude of 5 y. The third Pi2 that began at 06: 37 UT had a 150 set period and a somewhat larger amplitude (7 7). The oscillations observed on the satellite had an apparent period of 70-80 set and an amplitude of l-2 y as demonstrated in AB or the
C. C. LIN and L. J. CAHILL, JR.
I I DEC 24,1971
Fro. 2(a). MAGNETOGRAMS FORSUBsT7XW.¶ THAT OCCURREDON 24 DECEMBER 1971.
MKROPLILSATIONDATA A~~~~IATED WITH THE SuBsToRMsSHOWNIN Fro. 2(a). The onsets of the Pii events are shown at 05:56,06:22 and 06:38 UT.
A& component, while the power spectra of the third event (not shown) indicate that there were several periods present, a strong peak at 77 set and weaker peaks near 105 set, 52 set and 43 sec. Similar to the records at Sitka, the amplitude of the third Pi2 at the satellite was greater than that of the first two Pi2’s. The Pi2 that began at 05 : 56 UT might be associated with a small substorm apparent in the records at Fort Churchill, Great Whale River and in AB at the satellite. The second Pi2 at 06 : 22 UT could be interpreted as related to the onset of the expansive phase for a second substorm (Fort Churchill and GWR) and the third Pi2 that began at 06:37 UT was associated with another, larger, intensification of the current system. This is similar to Pi2 events associated with a trigger bay and main bay reported by Rostoker (1968). He stated that Pi2 oscillations often occur as a pair separated by 10-30 min. Kisabeth and Rostoker (1971) also found that accompanying the intensification of the current system there is enhancement of Pi2 activity 10-20 min after the Pi2 onset. The rapid run magnetograms from College showed similar Pi2 oscillations mainly in D and H components with a much smaller amplitude for the Pi2 that began at 05: 56 UT. The period for the second Pi2 was approximately 60-70 set but the period for the third
IN THE MAGNETOSPHERE
Pi2 was 140-150 sec. These later two Pi2’s had very distinct onset times. The same events were also observed at Fredericksburg, Boulder and Tucson. 9 December 1971 (10:00 UT)
Pi2 micropulsations associated with a substorm on 9 December 1971 are shown in Fig. 3. A sudden decrease in AB at the satellite began at 10: 00 UT and large magnetic bays were observed at College and Fort Churchill without distinct onsets. The Pi2 recordindicated that the substorm onset probably occurred at 1O:OOUT. The Pi2 micropulsations identified on the normal magnetogram from Honolulu, began at 08 :40 UT and ended at 11: 00 UT with a distinct set of pulsations beginning at 10: 00 UT. High time resolution magnetic data from Explorer 45 showed some fluctuations starting at ho8:40 UT, but there was no distinct onset at that time. Figure 3 shows there were three series of micropulsations beginning at 09: 17, 09: 36 and 10:00 UT. The first two series appeared in all. three components with a period of 90 set and amplitude about 2 y for the satellite datat The set that started at IO:00 UT had a larger amplitude of 49 y in all three components. Power spectra (not shown) indicate a strong peak at 80 set with several weaker peaks from 55 to 200 sec. In contrast to the results obtained from Explorer 45, the records from Sitka show an oscillation of period ~160 set and amplitude of 5 y that began at 09: 36 UT and lasted for 6 cycles. At 10:00 UT, shorter period Pi2 oscillations could be identified. The period was about 110 set and amplitude ~5 y. At College, which was about 01: 00 MLT to the west of Explorer 45 (L = 5.2), oscillations with periods 240 and 120 set were superimposed on higher frequency components and were observed in all three components beginning at about 08 : 40 UT. PC4 oscillations were identified at Fredericksburg from 08X)0 to 11: 00 UT. In view of these observations the first two series may not be Pi2 oscillations. Small perturbations can be seen in the ground magnetograms near 09 : 20 and 09 : 40 UT, but they do not resemble substorm onsets.
FIG. 3. MICROPULSATIONDATA ASSOCIATED WEH A SUBSTORMON 9 DECEMBER1971.
A large amplitude wave event began at 10:00 UT.
C. C. LIN and L. J. CAHILL, JR.
16 February 1972 (02:35 UT) The ground magnetic records shown in Fig. 4(a) indicated that a small substorm was observed at stations close to midnight, Fort Churchill, Baker Lake and Great Whale River. There was virtually no substorm activity at College. Although none of the nine stations we have consulted identified Pi2 oscillations during the time interval of interest, the small magnetic field activity at the satellite at this time made it easier to identify apparent Pi2 oscillations. Oscillations of amplitude 2 y started at 02: 35 UT, standing out distinctly as can be seen from Figs. 4(b) and (c). The ground records at Fredericksburg and Great Whale River showed the onset of the wave clearly at 02: 35 UT. The amplitude at Fredericksburg was approximately 5 y in H and D components and the period about 70 set, while at Great Whale River the amplitude reached 10 y, 5 y and 25 y for H, D and Z components respectively and the period was about 120 sec. Except for the D component, the fluctuations at Great Whale River were quite distinct. A component with period near 50 set was also discernible in the original records. Power spectra for the satellite data, shown in Fig. 4(d), indicate that there were four peaks with periods of 130,70,55 and 42 set, dominated by the peak at 130 sec. Some of these periods were observed on the ground. The polarization of the wave seen on the satellite was anticlockwise for the transverse portion of the peak near 130 sec. Pi2 oscillations were also observed at Boulder, Tucson and San Juan, but not at College and Sitka in the late afternoon at this time. The Pi2 oscillations were the only evidence of the substorm at Explorer 45 until the increase in AB after 03 : 00 UT.
FEB. 16.1972 I .3w co
w I 1
Fro. 4(a). MA~ETOGRAMSFOR
17 December 1971 (04:16 UT) The substorm records are shown in Figs. 5(a) and (b) and the micropulsation activity is presented in Fig. 5(c). The apparent onset of the expansion phase was approximately
IN THE MAGNETOSPHERE
L 4.75 HLAT 10.3
o- ----._ -10.
FIG. 4(b). MAGNETIC RECORDS FROM lhPLo=R 45 (ORBIT 288) FOR THE SAME EVENT SHOWN IN Fm.4(a). Note expanded AB scale compared to
FIG. 4(c). MICROP~~ONDATA~IATJZD WITH THE SUBSTORM SHOWN FIGS.~(~)AND (b).
The onset of the Pi2 activity is shown at 02:35 UT.
Fig. l(b). PERIOD fsec) I
?i Cr z
i-! FIG. 4(d). POWER
SPECTRA FOR THE INTERVAL 02:35-02:53 I_JTFOR ExRLORRR PULSATION DATA SHOWN IN FIG. 4(c).
C. C. LIN and L. J. CAHILL, JR.
at 04: 16 UT as determined from the stations close to local midnight, San Juan, Great Whale River and Fort Churchill. The changes in satellite field strength AB and Inclination indicate that the field configuration was tail-like before the onset of expansion and dipolelike after the onset of expansion phase. The commencement of Pi2 oscillations recorded on Explorer 45 was approximately at 04: 16 UT and the same event was observed on the ground at Boulder with no apparent time delay. Predominantly 75 set but also small 55 and 100 set oscillations were observed with Explorer 45 (Figs. 5(c) and (d)). The polarization for the peak near 75 set was anticlockwise. Hodographs for the 75 set peak are shown in Fig. 5(e). The peak to peak amplitude was 2 y for all three components of the 75 set peak. At Boulder, about 0l:OO MLT to the west of the meridian through Explorer 45, the oscillation period was 50 set and the peak to peak amplitude was 5 y in Hand D components and 4 y in the Z component. They were damped rapidly. To the east of the satellite meridian at a low latitude station, San Juan, the observed oscillations had a period about the same as that at Boulder but the amplitude was about 3.5 y in the H component. The D and Z components had a much smaller amplitude. The Pi2 event was also observed on the ground at Honolulu, Tucson and Sitka. They all exhibited a similar wave form to that observed at Boulder. Prior to
I 1.9 I
I I I
Gl I50 H(
0. -15. I
FIG. 5(a).MAGN~T~GRAM~ FOR A SUBSTORM THAT 00 CURREDON 17 DECEMBER
FIG. 5(b). MAGNETIC RECORDS FROMEXPL~RRR ~~(ORBIT 99) FORTHE SAME EVENTSHOWNIN FIG. S(a).
IN THE MAGNETOSPHERE
147 # I q::“~;..:_J\‘L”‘:lr:J?--141 1
DATAAWXLWED WITHTHESUBSTORM SHOWN IN FIGS.5(a) AND(b). FIG. S(c). MICROPUL~ATION The onset of the Pi2 activity is shown at 04: 16 UT. PERIOD 200
Wp3445 DEC.l7;;1971 I A% /'\ -
I IO FREQUENCY
FIG. 5(d). POWERuwx~~
PORTHEINTERVAL 04: 16-04: 33 UT FOREXPLORER 45 MICROPULSATION DATA SHOWNIN FIG. 5(c).
the Pi2 oscillation, beginning at 03 : 38 UT, there were sinusoidal oscillations at the satellite of period -80 set and amplitude l-5 y. These oscillations were also observed at San Juan and Boulder. 22 November 1971 (06:35 UT) As shown in Fig. 6(a), the magnetic field data from the satellite indicated a tail-like to dipolar change in field configuration during the substorm. The ground records showed that
C. C. LIN and L. J. CAHILL, JR. FILTER : 65 -88
FIG.~(~). HOD~GRAMSFORTHEINTERVAL~~: 16-04:20UT FOREXPLORER DATA SHOWN IN FIG. 5(c).
the onset time for the event was approximately 06: 35 UT. Figure 6(b) shows that the micropulsations started at 06: 34 UT at Boulder and Tucson. Honolulu and Sitka identified a SSC at 06:35 UT and oscillations were identified at Honolulu as Pi2 beginning at 06: 35 UT and ending at 09: 10 UT with a remarkable individual event beginning at 06: 35 UT and ending at 07:02 UT. Therefore, the observed oscillations may be associated with the storm’s sudden commencement. The period of the oscillations was about 90-100 set and the amplitude was about 5 y in the H and D components at Boulder. A comparable period and amplitude were seen at Tucson, which was 0l:OO MLT to the east of satellite meridian. Due to a small error in satellite data tape time we were not able to produce an accurate vector field. Only the field magnitude AB is presented here in Fig. 6(b). It appears that the event at the satellite began at 06:36 UT, about 2 min later than observed on the ground. However, the ground records revealed that the amplitude increased at 06: 36 UT L
100 : 00
: _______ I
I 0600 2227
FIG. 6(a). MAGNETIC
I, 0700 2326
REXXRDS FROM EXPLORES 45 (ORBIT 22).
Pi2 PULSATIONS IN THE MAGNETOSPHERE I
I I I
.,,,.....- . ... .. . . .., _
-/. , ,,
.. .,. ‘.,_.
... ‘.... .>\ ,.. ...” -..,,; ..
FIG. 6(b). MICROPULSATIONDATA ASSOCIATEDWITH THE smmoRM SHOWN IN FIG. 6(a).
The wave event began at 06:34 UT.
and there were small oscillations between 06: 34 and 06: 36 UT in the satellite record. The satellite oscillations initially had a period of -80 set and suddenly changed to -180 set at 06: 39 UT, lasted for 3 cycles and then changed back to 80 set again. The peak to peak amplitude was about 4 y, somewhat smaller than observed at Tucson and Boulder. The same event was also observed at Sitka, San Juan and Honolulu beginning at -06 : 34 UT. The rapid run magnetograms from College show a violent field perturbation starting at 46: 36 UT. 15 Junuury 1972 (12:50 UT.) Ground and satellite magnetograms, not shown, indicated that a substorm expansion occurred near 12: 50 UT. Micropulsation activity during the period of interest is shown in in Fig. 7. For the same reason as in the previous event, we show only the magnitude, AB There were also data gaps in this interval. The commencement of Pi2 observed at Honolulu was at 12:49 UT with an amplitude in Hand D of 2 y (hardly visible in the rapid run data shown) and with periods of 80 and 50 sec. The oscillations seen at the satellite with period 160 set and amplitude of 6 y began at 12: 52 UT, lasted for only 2 or 3 cycles and were followed immediately by smaller amplitude and shorter period oscillations. The records from the station at Guam show similar oscillations to those observed at Honolulu. The characteristics of these pulsation events observed in the magnetosphere and on the ground are summarizedin Table 2. The periods for the satellite dam are those of the dominant power spectral peaks and the periods and amplitudes (peak to peak) for the ground data are the apparent periods and amplitudes estimated from the first few cycles. The polarizations indicated as anticlockwise are as observed looking in the direction of the magnetic field line. 3. DISCUSSION
Periods are difficult to determine from even rapid run magnetograms so the apparent periods of the tist few cycles of the strongest signals of the ground records were chosen to
C. C. LIN and L. J. CAHILL, JR, I
i I 1250
I JAN 15.1972
FIG. 7. MICROPULSATIONDATA ASSOCIATBD WITH A SUBSTORMON 1s JANUARY 1972.
compare with the periods of dominant peaks of the satellite spectra. From Table 2, we see that the periods observed in space are somewhat different from those observed on the ground. In only two cases, 9 December and 16 February, are the estimated ground periods within the range of periods from the satellite spectra. In one other case, 24 December, the estimated ground period is well above the satellite period range. The ground period is at the lower edge of the satellite range on 17 December and at the upper edge on 4 January. On 15 January the ground period is half the satellite period. On 22 November, the estimated period of the oscillations observed on the ground is between the two periods estimated at the satellite. These observations may be consistent with the complex spectra observed by other workers in ground measurements. Here we only estimate apparent periods of the largest ground signals. If the amplitudes of major satellite signals are compared with the amplitudes of the first few cycles of the ground data, Table 2 shows that the amplitudes of the equatorial Pi2 oscillations detected by the satellite are smaller than those observed at Great Whale River, Fredericksburg, Sitka, Boulder and Tucson. The only exception is the 15 January event, when the amplitude at Honolulu is smaller than that at the satellite. It is generally agreed that Pi2’s observed on the ground are elliptically polarized in the horizontal plane in mid- and low-latitudes while aurora1 zone events usually have a strong vertical component. Evening sector ground observations between 54 and 65” (L = 3-5) indicate left hand polarization (anticlockwise) (Saito, 1969). For most of the events we have studied, it appears that the waves in space are strongly polarized in the direction of the magnetic field with a weak or absent transverse component. In the five events for which vector fields are available, one event is nearly completely compressional, and the sense of rotation for the weak transverse component of three of the others is anticlockwise as viewed northward along the main field direction. The 17 December event, when all components were about equal, was also anticlockwise. Thus the principally compressional waves in the magnetosphere are seen polarized in the H-D plane on the ground. Lanzerotti and Tartaglia (1972) reported a wave event (period = 104 set) which was detected by ATS 1 at 6.6 earth radii at the equator as a pure compressional wave while it was observed at College as a nearly pure transverse wave. Ground Pi2 oscillations, in general, contain several period components and the components in the spectra are usually not harmonically related. Recently, Roth and Orr (1973) presented several examples of harmonic spectra of Pi2 oscillations with a fundamental
IN THE MAGNETOSP~
period of about 150 sec. Our results indicate that the number of period components ranges from two to four. Although peaks at about 40-55 set and at 75-100 set are seen in the same event, there is no strong indication of harmonically related power spectra. A review of Pi2 generation theories was given by Doobov and Mainstone (1973b). The initial stage of Pi2 oscillations is generally considered to be the result of the impact of a stream of energetic particles on the closed field lines in the night side magnetosphere. The stream of energetic particles is thought to be formed due to the energy release from field line reconnection in the neutral sheet. From here on the theories vary. Some authors attribute the Pi2 oscillation to torsional oscillations of the field lines (Rostoker, 1967; Sakurai, 1970). A very interesting argument was presented by Rostoker, who suggested that the impact of a stream of energetic particles will cause the aurora1 zone field lines to oscillate in a torsional mode. The compressional waves we observe near L = 5 are in direct conflict with this model. It is possible, however, that the same event may take the form of a torsional oscillation in some other L shell. Saito and Matsushita (1968) proposed that pi2 results from poloidal oscillations of the closed magnetosphere and that the waves contain only one period. Our waves are more poloidal than torsional but usually contain more than one period. In addition, our results indicate that the waves are not completely polarized in the meridian plane. A small AB, occurs even for meridian plane oscillations when the satellite is not exactly on the equator, but the AB, component and clear elliptical polarization in some cases indicate torsional oscillations mixed with poloidal. Raspopov (1968), suggested that there are three possible ma~etosphe~c cavities where resonance conditions may exist for hydroma~eti~ waves propagating across the magnetic line of force. Instead of three cavities, Doobov and Mainstone (19735) proposed that, because of discontinuities in the magnetosphere, there are at least six cavities in the night side magnetosphere (see Fig. 2 of Doobov and Mainstone, 1973b). These authors find for the six cavities TX= 10,Ta= 33, T3 = 44, T4= 43, Ts= 77, T,= 87 sec. It is interesting to note that under typical magnetospheric conditions, oscillations in the Pi2 range near 40 and 80 set are predicted. As mentioned previously, periods near 45 and near 80 set occur in the satellite observations. Since the Pi2 oscillations are supposed to be due to the impact of energetic particles on the closed magnetosphere, we migh expect that cavities with the closed magnetosphere as boundary are the best candidates (T,= 44,T6= 77, TB = 87). In the events we examined, except 15 January, a period near 80 set is evident. In several events there is a weak peak near 45 sec. The larger periods over 100 set are not directly predicted but the model parameters could probably be adjusted to fit periods near 120 sec. In contrast to our results, Doobov and Mainstone (f973a), did not find a particular frequency which occurs more often than others. It should also be noted that the results of Doobov (1972) were based on the data from three rather low latitude stations. These results do not exhibit a latitude dependent amplitude, while some authors using aurora1 and sub-aurora1 zone data have observed the amplitude of Pi2’s to be latitude dependent (Jacobs and Sinno, 1960). Cladis (1971) has proposed a field line resonance in which gradient drift of energetic particles with a sharp azimuthal density front (such as might accompany a substorm) can cause the field lines to oscillate as standing waves in the meridian plane. This agrees with the strong ABe component we observe. His model indicates periods between 120 set and 420 set (Pi2 and PCS pulsations). The lower end of this range is appropriate for the longer periods that we observe. 10
C. C. LIN and L. J. CAHILL,
In summary we have observed seven evening quadrant pulsation events with near simultaneous onsets at Explorer 45 and on the ground. Six of these were identified as Pi2 by ground observers. These distinct onset events occurred near the onsets of the expansion phases of magnetospheric substorms and this suggests that the substorm onset is the time of generation of some Pi2 pulsations. Pi2 pulsations may be a useful tool to establish the onset of substorms throughout the evening quadrant magnetosphere. Note that one Pi2 event (16 February) was observed at 18 : 50 MLT and one at 23 : 00 MLT (22 November). The observations do not agree with all of the theoretical predictions but some of the comparisons are promising. In particular it will be useful to compare these events with the proton fluxes measured on this same satellite to search for evidence of Cladis’ mechanism. Acknowledgements-This work was supported by the National Aeronautics and Space Administration under Grant NAS s-11173. Magnetograms used in this study were obtained from World Data Centre A. We have benefited from the assistance of Bodo Parady, Yue Lee and Richard Madden of the University of Minnesota. AKASOPU,S.-I. (1964). The development of the aurora1 substorm. Planet. Space Sci. 12,273. CAHILL, L. J., JR. (1973). Magnetic storm inflation in the evening sector. J.geophys. Res. 78,4724. CAIN, J. C. and S-Y, R. E. (1970). Magnetic field mapping of the inner magnetosphere. J. geophys. Res; 75.4360. CAMPBELL,W. H. (1967). Geomagnetic pulsations. In Physics ofGeomagnetic Phenomena (Eds. S. Matsushita and W. H. Camobell). Academic Press. New York. CHOY, L. W., AWO~DY, k. L., POTTER, %‘., KINTNER, P. and CAHILL, L. J. JR. (1971). Fieldparticle currents near an auroral arc. J. geophys. Res. 76,8279. CLADIS, 3. B. (1971). Multipl coupled oscillation of field fines in the magnetosphere: Modulation of trapped particles and ionosp IIeric currents. f. geophys. Res. 76, 2345. CQRONITI, F. V., MCPHERRON,R. L. and PARKS,G. K. (1968). Studies of the magnetospheric substorm-3. Concept of the magnetospheric substorm and its relation to electron precipitation and micropulsations. J.geophys. Res. 73,1715. DEFOE, S. E. and MCILWAEN, C. E. (1971). Plasma clouds in the magnetosphere. J. geophys. Res. 76, 3587. Dooaov, A. L. (1972). Latitude effects on the amplitude of Pi2 micropulsations. J. atmos. terr. Phys. 34, 1945. Frequency spectra Dooaov, A. L. and MAINSTONE,J. S. (1973a). Investigations of Pi2 micropulsations-I. and polarization. Planet. Space Sci. 21,721. DOOBOV, A. L. and MAX-NE, J. S. (1973b). Investigations of Pi2 mi~op~sations-H. Relevance of observations to generation theories. Planet. Space Sci. 21,731. DWARKIN, M. L., ZMUDAA. J. and RADFORD,W. E. (1971). Hydromagnetic waves at 6.25 earth radii with periods between 3 and 240 seconds. J. geophys. Res. 76,3668. InPhysics and Chemistry in Space (Eds. J. G. Roederer JACOBS, S. A. (1970). Geomagnetic micropulsations. and J. Zahringcr) Vol. 1. Springer, New York. JAMBS, J. A., KATO, Y., MAXWX-IITA, S. and TROIT~KAYA.V. A. (1964). Classification of geomagnetic micropulsations. J. geophys. Res. 69, 180. JAFe$$3J.3_$and SINNO, K. (1960). World-widecharacteristics of geomagnetic micropulsations. J.geophys.
J. L. and ROSMKER, G. (1971). Development of the polar electrojet during magnetic substorms. J. geopi~ys. Res. 76,681X LANZEROTTI, L. J. and TARTACSLIA, N. A. (1972). Propagation of a magnetospheric compressional wave to the ground. J. geophys. Res. 77, 1934. LEZNIAK, T. W. and WINCKLERJ. R. (1970). Experimental study of macnetosnhere motions and the accele&tion of energetic electrons during substor&. J. geophys. 8es. 75,707s. * MCPHERRON,R, L. (1973). Satellite studies of magnetospheric substorms on August 15, 1968-l. State of the magnetosphere. J. geophys. Res. 78, 3044. RASPOPOV,0. M. (1968). Possible excitation mechanism of type Pi2 geomagnetic field pulsations. Geomngn. Aeron. 8,257. RO~TOKER,G. (1967). The frequency spectrum of Pi2 ~~opul~tion activity and its relationship to planetary magnetic activity. J. geophys. Res. 72, 2032.
IN THE MAGNETOSPHERE
ROSTOKER,G. (1968). Macrostructure of geomagnetic bays. J. geophys. Res. 73,4217. ROSTOKER,G. (1972). Polar magnetc substorms. Rev. Geophys. Space Phys. 10, 157. ROTH, B. and ORR, D. (1973). Harmonics in spectra of Pi2 pulsations. Planet. Space Sci. 21, 1273. SAITO, T. (1961). Oscillation of geomagnetic field with the progress of Pt-type pulsation. Sci. Rep. Tohoku Univ. 5, Geophys. 13,53. SAITO, T. (1969). Geomagnetic pulsations. Space Sci. Rev. 10, 319. S~rro, T. and MATSUSHITA,S. (1968). Solar cycle effects on geomagnetic Pi2 pulsations. J. geophys. Rec. 73,267. SAKUIXAI,T. (1970). Polarization characteristics of geomagnetic Pi2 micropulsations. Sci. Rep. Tohoku Univ. 20, 107. S~unnr, B. P. (1973). On the occurrence of Pi2 micropulsations. Planet. Space Sci. 21,831. WHALEN, B. A. and MCDIARMID, I. B. (1972). Observations of magnetic-field-aligned auroral-electron precipitation. J. geophys. Res. 77, 191.