Satellite Studies of Magnetospheric Substorms on August 15, 1968

Transcription

1 VOL. 78, NO. 16 JOURNAL OF GEOPHYSICAL RESEARCH JUNE 1, 1973 Satellite Studies of Magnetospheric Substorms on August 15, Ogo 5 Magnetic Field Observations R. L. MCPHERRON, M.P. AUBRY, 2 C. T. RUSSELL, 3 AND P. J. COLEMAN, JR. In this paper we examine the changes in the magnetic field on the midnight meridian 15-8 R r behind the earth. These changes are conveniently divided into two main phases: a growth phase and an expansion phase. The beginnings of the growth phase coincide within experimental error to the arrival of a southward solar wind magnetic field at the dayside magnetopause and also to the beginning of the growth phase at synchronous orbit and ground observatories. Durihg the growth phase the magnitude of'the field increases the lobe of the tail. In the cusp both the magnitude and the inclination of the field increase. Slow changes in the cross-tail field accompany these effects. Nearly coincident with the onset of the polar substorm the lobe field magnitude begins to decrease, the cusp field magnitude and inclination decrease, and there are sudden changes in the cross-tail field. About his time the plasma sheet begins to expand. Inside the expanding plasma sheet and associated with the field rotation are large fluctuations in the field. The response of the field at 8 R r was simultaneous with the ground onset, whereas further back at 11.3 and 2.4 R above the expected neutral sheet the largest changes were delayed by 15 min. In the preceding papers of this series we have used data obtained from a number of sources to study several magnetospheric substorms on August 15, This paper continues this study with a detailed examination of the substorm changes in the magnetic field in the midnight meridian 15-8 R behind the earth. These observations were made by the UCLA flux-gate magnetometer [Aubry et d., 1971] on the 0go 5 spacecraft as it was inbound, down out of the north lobe of the tail. This and subsequent papers of this series are limited to observations made by various instruments on Ogo 5. To aid understanding, the magnetic field observations of this paper are presented in three different ways. First, the three components and Department of Planetary and Space Science and Institute of Geophysics and Planetary Physics, University of California, Los Angeles, California Department of Planetary and Space Science, University of California, Los Angeles, California 90024, and ESRO-NASA University Research Associate on leave from Groupe de Recherches Ionosph riques du C.N.R.S., Paris, France. 8Institute of Geophysics and Planetary Physics, University of California, Los Angeles, California Copyright 1973 by the American Geophysical Union. the magnitude of the field are discussed in geocentric solar magnetospheric coordinates (GSM). Second, the same observations are repeated in terms of parameters related to the dipole field. These parameters include the deviation of the field magnitude from that of a dipole, the angles of inclination, and the rms deviation of the field. Finally, the vector field observations are projected into the noonmidnight meridian plane. These projections are compared to field lines of the dipole and measured quiet time magnetic field. Our results indicate that significant changes occur in the near tail during a magnetospheric substorm. We find that during substorms there are very large. changes in the magnetic field configuration in this region. These changes are systematically related to the onset of the expan- sion phase identified with ground magneto- grams. For two worldwide substorms early on August 15, 1968, our results suggesthere were two main phases, a growth phase and an ex- pansion phase. During the growth phase the tail field magnitude increases, and its orientation becomes more taillike. In the expansion phase the magnitude decreases, and the orientation becomes more dipolar. Accompanying the changes in the expansion phase are largeamplitude fluctuations in the field. About the 3068

2 MCPHERRON ET AL.' SUBSTORM STUDIES, onset of the expansion phase there are also ported by Sugiura et al. [1968, p. 6708]. They large variations in the cross-tail component of characterized the near-tail substorm behavior the field, suggesting changes in field-aligned as a '... collapse of the tail field configuration currents. and a return toward the dipole configuration Since the observations presented below are in the tail sector that is located near the melimited to two substorms, it is important to ridtan plane of the bay onset.' These authors establish the substorm sequence as well as pos- again interpreted the observations as evidence sible by using previous reports. Consequently, of a near-earth origin for the substorm disthe following section is devoted to a brief re- turbance. view and summary of tail field observations. In Substorm observations of the vector magthis review we limit ourselves to. investigations netic field at the synchronous equatorial orbit that have dealt with substorm changes in the of ATS 1 have been reported by Cummings field configuration. As we will show, our obser- and Coleman [1968], Cummings et al. [1968], vations are in general agreement with these Coleman and McPherro.n [1970], and Mcreports, provided our interpretation of the Pherron and Coleman [1970]. These authors ground magnetograms is correct. In this. paper find that substorm effects are most pronounced our presentation is based on conclusions ob- in the premidnight sector. Prior to the onset tatned in papers 1 and 2, regarding the onset of the expansion phase the field becomes deof substorm expansions. pressed and slightly inclined away from the earth. During the expansion phase the field REVIEW OF TAIL MAGNETIC FIELD OBSEaV T ONS recovers. This recovery is accompanied by largeamplitude field fluctuations and also by occa- The possibility that the tail magnetic field sional azimuthal field variation possibly due to is fundamentally involved in substorms has field-aligned currents. Cummings ei al. [1968] been considered for some time on a basis of interpreted their observations in terms of a indirect evidence. For example, A tkinso near-earth model of the substorm disturbance. [1966], Ax/ord [1967], and Dungey [1968] They sugges that a partial ring current inoutline models in which solar wind energy is jected immediately after the expansion phase stored in the tail as magnetic field energy and of one substorm is responsible for the depresreleased through magnetic merging. All these sion phase of the next. Loss of this partial ring models have emphasized the outer tail region is then the cause of the recovery. as the source of the substorm disturbance. Camidge and Rostoker [1970] have used In contrast to the models, Heppner et al. vector field observations on Imp I and 2 in the [1967] suggested that the substorm disturbance more distant tail (~25 R, to support the is a consequence of plasma convection in the tail field merging model of substorms. They find tail toward the earth and that the initial dis- substorm effects on the tail coincident with --+6 turbance or triggering is located close to the min of the substorm onset. The effects were earth. Using Ogo I observations of the magnetic generally short lived at large distance and field in the near tail, these authors found sig- sluggish close in. The predominant effect was nificant changes in the field magnitude asso- a decrease in magnitude primarily due to a large ciated with substorms. The changes were either decrease in the GSM X component. Also evian increase or a decrease in field magnitude, dent was a strong rotational effect causing a depending on whether the satellite was in a transition from a taillike to a more dipolar region of the tail in which the field magnitude field. A peak in the magnitude of these effects was less than or greater than that of a dipole. at 25 R and strong negative Z components The delay of these changes relative to the on- were interpreted as evidence of magnetic field set of substorms as well as a number of morpho- merging. logical considerations were used to support their Imp 4 magnetic field ob rvations reported near-earth model. by Fairfield and Ness [1970] have also been Vector field observations of the near-tail re- used to support the magnetic field merging gion made on Ogo 3 during substorms were re- model. These authors find that in an early

3 3070 McPHERRON ET AL.: SUBSTORM STUDIES, 4 phase of substorms the tail field magnitude internal magnetic field. The eroded flux is transincreases, the g component decreases, and the ported by the solar wind and added to the geoplasma sheethins. Later in the substorm these magnetic tail. The magnitude of the field begins effects are reversed. The fact that more field to increase in the lobe of the tail, and the plasma lines cross the equator after a substorm expansion than before indicates that the tail is in a sheet thins. In the nighttime cusp region the field magnitude begins to increase and becomes lower-energy state at the end of a substorm. more and more taillike. Closer to the earth at This suggests that substorm energy is initially synchronous orbit the field magnitude decreases. stored in the tail field. Merging of field lines Provided that the solar wind field remains is one po ible way of releasing this energy. Evidence that the tail field increase during southward, these changes eventually reverse suddenly. At midnight, coincident with the onsubstorms is a consequence of erosion of the set of the auroral and polar magnetic substorms, dayside of the magnetosphere has been pre- the field at synchronous orbit begins to recover sented by Aubry et al. [1970]. These authors in magnitude. In the cusp region, possibly with found that a 2-RE inward motion of the mag- some delay, the field magnitude begins to denetopause followed the onset of a southward crease and also to rotate toward a more dipolar component of the solar wind magnetic field, even though there was no significant change in configuration. At the same time the plasma sheet begins to the solar wind momentum flux. The inward expand, and the lobe field magnitude to demotion produced no compression of the dayside magnetosphere but did cause the distant tail field magnitude to increase. The increase was followed by a substorm observed both in the auroral zone and in the tail. Russell et al. [1971] have reported Ogo 5 magnetic field observations in the tail during weak substorm activity. Their results generally confirm those discussed above. The authors note, in addition to previously reported features, that large magnetic fluctuations and sudden changes in declination are associated with the sudden crease. Large-amplitude fluctuations and rapid changes in the cross-tail component of field occur during this phase. The Ogo 5 observations, presented below, are in agreement with these reports. Taken together with the measurements of the solar wind magnetic field, the field at synchronous orbit, and the ground observations, they provide the most complete set of observations of the substorm sequence yet reported. ORBIT Or OGO 5 ON AUGUST 15, 1968 changes in field magnitude and inclination. A separation of the effects on the tail due to a southward component of the solar wind On August 15, 1968, the orbit of Ogo 5 was ideally situated to study events in the geomagnetic tail. Projections of this orbit in the three magnetic field and magnetospheric substorms orthogonal planes of GSM coordinates are has been carried out by Aubry and McPherron [1971]. They find in the main lobe of the tail the magnetic field increases slowly when he shown in Figure 1. The equatorial projection in the lower left panel shows that, at the beginning of August 15, Ogo 5 was about 15 RE solar wind field turns southward and decreases behind the earth and nearly in the midnight slowly after the onset of a substorm expansion. meridian. The satellite was inbound and by The plasma sheet thins when the field turns southward and expands when it turns northward. A substorm expansion also causes the plasma sheet to expand. From the preceding reports we arrive at the following summary of the substorm sequence of changes in the magnetic field. A southward component of magnetic field reaches the magnetopause in the solar wind. Erosion of the dayside magnetosphere begins to move the magnetopause earthward without compressing the 0900 UT was well into the inner magnetosphere, passing through perigee at 1034 UT. The solar meridian plane projection in the upper left panel indicates that the satellite was coming down out of the northern lobe of the tail, passing through the ecliptic plane at about 0900 UT. The dawn-dusk plane projection in the upper right panel emphasizes the midnight meridian alignment during the first half of August 15. A detailed listing of essential orbit information is given in Table 1.

4 MCPHERRON ET AL.' SUBSTORM STUDIES, SOLAR MERIDIAN PLANE PROJECTION GSM Z vs X DAWN-DUSK PLANE PROJECTION GSM Z vs Y AUG, 15 z 15 AUG I , I0-5 X --'- l'uøh I I0 15 MAGNETOSPHERIC EQUATORIAL PROJECTION GSM Y v$ X,2 8 I Y 12 I 16 L AUG I ', t I --'"'-..._ I i X,, _-- TO " " ' V 5 SUN Fig. 1. Orbit 63 of Ogo 5 during August Satellite positions at 4-hour intervals are shown by tick marks along the three projections o.f the trajectory in GSM coordinates. During the first half of August 15, Ogo 5 was inbound, down out of the north lobe of the tail. OGO 5 MAGNETIC FIELD OBSERVATIONS Presentatiem in GSM co.ordinates. The three components and magnitude of the magnetic field measured by the UCLA magnetometer on August 15, 1968, are plotted as 1-min averages in Figure 2. Vertical dashed lines are the onsets of substorm expansion identified in paper 1. Vertical arrows near the bottom of the figure refer to times of significant events at Ogo 5. General features of substorms in the near tail reported previously by Russell et al. [1971] are clearly evident for the 0430 and 0714 UT substorm expansions. These features include a relatively slow increase in field magnitude followed by a more rapid decrease. Accompanyiag these changes are a slow decrease in the g component followed by a rapid increase. About the beginning of the sudden decrease in field mag- nitude there were sudden changes in lhe Y component. An examination of the timing of the changes in field magnitude relative to the onset of substorm expansion as determined in paper 1 reveals an important fact: there is a close corre- spondence between the onset of the substorm expansions and the beginning of the decrease in field magnitude. For the 0430 UT substorm the precise beginning of this decrease is not clear, since there are two maximums. The most likely times are either 0418 or 0428 UT. The latter time is the beginning of a larger and more

5 3072 McP ERRON ST AL.' SUBSTORM STUDIES, 4 TABLE 1. Ogo 5 Orbit Information on August 15, 1968 GSM Location, Rr Time, ' UT X Y Z Radial Z above Magnetic Dipole Distance, Local L Neutral Latitudes, Tilt, Rr Time Parameter Sheet, Rr deg deg * * I I -1 3 I ' I 1 O8O I O I * Onset of substorm expansion. sudden drop. For the 0714 UT substorm the the large changes in BZ occur near the onset beginning of the decrease coincides with the of the substorm expansion. Third, the beginning onset of the expansion phase within I min. of the increase in BZ also appears to be asso- Examination of the timing of the changes in ciated with the onset of the expansion phase. the field components also shows several impor- For the 0714 UT expansion these times are rant facts. First, the changes in magnitude are simultaneous, but for the 0430 UT expansion primarily a result of changes in BX. Second, they are not precisely simultaneous, and, in i 3,._..,--,--.,--- o3 40 (,.9, 20 : o 20 m o n 60 8o N O 40 8O 60 o 40 D 20.i.. ;! ; ; : ;...,... : 0 : : : j ß -]....,4 J ' ' : ::,, : : : : :,,!, i!! i i i i i' i i i i i i i : ' 0 i i i i i UNIVERSAL TIME )0 Fig. 2. One-minute averages of the magnetic field in GSM coordinates made by UCLA magnetometer on Ogo 5 between 0000 and 0900 UT, August 15, The onsets of the substorm expansions are shown by vertical dashed lines. The times of significant field changes are shown by vertical arrows.

6 fact, the beginning of the slow increase in BZ could have been either 0413 or 0425 UT. MCPHERRON ET AL.: SUBSTORM STUDIES, 4 A comparison of the g'component of the field between the 0430 and 0714 UT expansions shows a significant difference. Just before the 0430 UT expansion, BZ was very small and may have been negative. (The BZ base line is not known to better than a few gammas.) At this time, BX was larger than Consequently, the field orientation was essentially parallel to the neutral sheet. Such an orientation is typical of the lobe of the tail. For the 0714 UT expansion, BZ was about half of BX, and thus a relatively large inclination of the field was indicated. This orientation is typical of the cusp, i.e., the transitional region between the taft and the inner more dipo]ar field. Conse- quently, we conclude that the satellite was in two distinct regions. of the earth's field, the lobe and the cusp, just before the onsets of the two substorm expansions. In radial distance, Table 1 indicates that the satellite was at 11.7 and are to be compared to 0330 and 0640 UT determined in paper 1 and to 0340 and 0630 UT 3073 determined in paper 2 as the beginning of the growth phases. Within the errors of our measurements these times are identical. Finally, we comment on the 0220 UT substorm expansion. As was discussed in papers 1 and 2, the effects of this substorm were relatively weak at the ground and also at synchronous orbit. Furthermore, the solar wind magnetic field was weakly and only intermittently southward. Examination of the tail field data in Figure 2 at this time reveals hardly any significant changes. From Table 1, however, we note that Ogo 5 was 14 R from the earth and 4.2 R above the expected position of the neutral sheet. Since this should be high up in the lobe of the tail, it is possible that a weak substorm might not have observable consequences. Presentation in dipole parameters. A different presentation of the UCLA magnetic field observations emphasizing the relation to a dipole field is given in Figure 3. The top panel shows the difference in magnitude between the observed field and a dipole field. The second 7.9 R. We pointed out here that both the 0430 and and third panels show the angles of inclination the 0714 UT expansions appear to have had two and declination of the field. Inclination is the stages. In the first expansion after the changes angle the field vector lies below a plane peraround 0430 UT, there was at 0444 UT very sudden increase in BZ and a corresponding decrease in BX. These changes indicate a rapid rotation of the field from a taillike to a more dipolaf field. Since the timing of the onset determined in paper 1 is not likely to be in error by 15 min, we conclude that this rotation was delayed relative to the onset of the expansion phase. In the next expansion there are two stages as well. In the first stage beginning at 0714 UT pendicular to a radius vector through the satellite. The dashed line indicates the expected inclination for a dipole field. Declination is the angle between the projections of the magnetic dipole and the field vector in this plane. If the field projection points east of the magnetic meridian, declination is positive. The fourth panel repeats the GSM Z component discussed With reference to Figure 2. The bottom panel shows 1-min averages of the rms fluctuations in the field for variations with periods of less the large changes in BZ and BX correspond to than 15 sec. a very rapid rotation of the field toward a more dipolar configuration. In the second stage a large increase in BZ alone results in an in- Studying the data plotted in the top panel, we find additional evidence that tlie satellite was in two distinct field regions. at the time of crease in field magnitude without a correspond- the two main expansions. At 0400 UT the sateling rotation. We turn next to the timing of field changes at lite was in a region in which the trend ins the field magnitude was constant and enhanced rela- Ogo 5 relative to the beginning of the substorm tive to a dipole, whereas at 0700 UT the trends growth pha s as determined in papers 1 and 2. We find in both cases a rather close correspondence. At 0320 and also at 0640 UT, BX began to increase and BZ to decrease. These times show the field magnitude was becoming depressed relative to a dipole. Comparing the changes in inclination of the field to the onset of the 0430 and 0714 UT expansions, we clearly see the field rotations discu ed in the preceding section. For both

7 3074 McPHERRON ET AL.' SUBSTORM STUDIES, 4 6O o 8O 6O O 2O o -60 4O 20 0 : rn o I 0220 I I I I I O UNIVERSAL Fig. 3. Same data as in Figure 2 using parameters related to the dipole field (see the text for definitions). TIME substorms the inclination increased during the onset, as was the field rotation. growth phase and decreased during the expan- An interesting feature of Figure 3 in contrast sion phase. The delay in the rotation for the to Figure 2 is that the effects of the 0220 UT 0430 UT sdbstorm is evident, as is also the lack expansion are more apparent. These include of rotation associated with the 0728 UT increase in BZ in the 0714 UT substorm. The declination plotted in the third panel of Figure 3 exaggerates the changes in the Y cornponent of the field. For both the 0430 and the a decrease in inclination, an increase in declination and BZ, and rms fluctuations. PRESENTATION IN NOON-MIDNIGHT MERIDIAN PLANE 0714 UT substorm the declination slowly de- The magnetic field observations of Figures 2 creased and changed sign during the growth and 3 are made clearer in Figure 4 by comparphase. During the expansion phase it rapidly ing them to the expected field line configuration increased. The beginning of this increase coin- in the noon-midnight meridian plane. Thin solid cided very closely to the onset times determined lines represent the field lines of a pure dipole in paper 1. field, whereas thin dashed lines represent the Rms deviations plotted in the bottom panel average magnetic field configuration reported show the existence of large-amplitude fluctua- by Fairfield [1968]. The magnetic equator and tions in the field during the substorm expan- expected neutral sheet [Russell a d Brody, sions. These fluctuations appear to be most 1967] at 0430 and 0714 UT are shown by closely linked to the increase in the Zcom- heavy dashed and solid lines near the negative ponent. The fluctuations in the 0430 UT sub- GSM-X axis. Since the dipole is so nearly storm were delayed relative to the expansion perpendicular to the earth-sun line at these

8 ZGSM IcPHERRON ET AL.' SUBSTORM STUDZSS, , VELA 0430! XGSM (TO SUN) / - o XG$ M Fig. 4. Noon-midnight meridian plane projection of esrth's magnetic field. Thin solid and thin dashed lines are dipole and measured [Fairfield, 1968] field lines. The neutral sheet positions at 0430 and 0714 UT are determined by the formuls of Russell and Brody [1967], the observed magnetic field is shown by solid arrows along the Ogo 5 trajectory. The dashed arrows at 0400 and 0700 UT are the dipole field for compsrison. times, it is unlikely that the neutral sheet position differs significantly from those shown. The probable locations of the upper and lower edges for a or ñ3-re-thick plasma sheet are drawn for 0430 UT near the 17-R circular orbits of the Vela satellites. The plasmasphere as observed by Ogo 5 at 0950 UT (paper 3) is indicated by shading inside L The projection of the observed magnetic field into the noon-midnight meridian plane is shown by vectors along the two segments of the satellite trajectory. The first segment between 0300 and 0500 UT displays every tenth 1-min average of the field (except for one vector at 0455 UT). At the beginning of the segment the. field was taillike but slightly less so than the average configuration of Fairfield [1968]. As the satellite moved in and down, the inclination increased until at 0400 UT it was nearly parallel to the neutral sheet. For comparison the dipole field at 0400 UT is shown by a dashed arrow. The field remained approximately parallel until 0444 UT, when it began to rotate toward a more dipolar configuration. This decrease was of short duration (see Figure 3) so that at 0450 UT the field vector was again temporarily parallel to the neutral sheet. After 0450 UT the rotation was very rapid, so that at 0455 UT the field was quite dipolar in orientation. From the position of Ogo RE above likely that the satellite was in the lobe of the tail rather than the plasma sheet. This suggestion follows from the report by Bame et al. [1967] that the half-thickness of the plasma sheet is 2-3 R. From 0300 to 0418 UT there were no decreases in field magnitude as might be expected if the satellite had entered the diamagnetic plasma sheet. Consequently, it must have remained in the lobe until at least 2.5 RE above the neutral sheet. Between 0418 and 0500 UT the field magnitude decreased in several steps. The sudden decrease in magnitude at 0444 UT, 15 min after the onset of the substorm expansion, is probably the diamagnetic effect of an entry of the plasma sheet cau d by the substorm expansion. The sudden field rotation and field fluctuations support this interpretation. If this is the case, the plasma sheet must have had a half-thickness of 2.3 RE at this time. If we accept 10 km/sec as the velocity of expansion of the boundary of the plasma sheet [Hones et al., 1971], we can estimate the probable half-thickness of the sheet 14 min earlier at 0430 UT. With these numbers we find 0.9 RE. Using a velocity twice or half as large gives values between 0 and 1.6 As will be shown in papers 5 and 6 [Kivelson et al., 1973; West et al., 1973] of this series the preceding interpretation is supported by the particle :observations on Ogo 5. Prior to 0432 the expected neutral sheet at 0300 UT it seems UT both electron and proton count rates were

9 3076 low or below detector thresholds. At 0432 UT, just after the onset of the expan on phase, an increase in the proton count rate was observed [West et al., 1973, Figure 1] (paper 6). Then at 0446 UT there was a several order of magnitude increase in count rates for both electrons and protons. The coincidence of. these large particle changes with the magnetic effects described above suggests that the satellite passed through a boundary into the plasma sheet. We interpret the appearance of weak fluxes of protons at 0432 UT in the absence of a simultane- ous appearance of electrons as being due to enhanced diffusion of protons from the plasma sheet at this time, perhaps owing to an increase in the ULF noise at the onset of the plasma sheet expansion [Russell et al., 1971]. The second segment between 0650 and 0730 UT displays every fifth 1-min average. As in the previous substorm the field at the beginning of the gment agreed very closely with the average magnetic field configuration of Fairfield [1968] and was distinctly taillike (compare to dipole field at 0700 UT). As the substorm progressed, the inclination increased, reaching a maximum at 0715 UT. Following the onset of the substorm expansion the field rotated rapidly and became nearly dipolar. On the basis of its spatial location Ogo 5 should have been well inside the plasma sheet at 0714 UT. In fact, it is possible that it was inside its inner boundary as defined by Vasyliunas [1968] in terms of low-energy electrons. If we assume that the inner edge of the tail current is located near 10 Rr, the satellite was inside it as well. However, the very taillike field at 0714 UT suggests that the inner edge of this current was closer to the earth than 10 R. The rapid increase in field magnitude is likely the result of passing out of the plasma sheet into the lobe of the tail rather than a conse- quence of a temporal change in the cross-taft current. The change from a field strength depressed relative to the dipole field to one enhanced relative to the dipole is consistent with this interpretation. For this to occur, the plasma sheet at 8 R must have thinned during the growth phase, reaching a minimum thickness of less than I Rr just before the onset of the expansion. The preceding suggestion is somewhat sur- MCPHERRON ET AL.: SUBSTORM STUDIES, 4 prising but is supported by particle observations on the same satellite. In paper 5 [Kivelson et al., 1973] it is shown that the electron fluxes rapidly decreased near the end of the growth phase, possibly owing to plasma sheet thinning. In paper 7 [Buck et al., 1973] it is shown that a large asymmetry in the east-west ratio of energetic protons can be explained in terms of a steep vertical gradient of the proton flux. Paper 7 also derives the velocity of thinning and the scale length of the boundary by using the energy dependence of this effect. Consequently, we conclude that the field signature observed near 0714 UT is due to a strong tail current and an exceedingly thin plasma sheet at 8 Rr. The second stage of the 0714 UT substorm expansion was characterized by a rapid increase in field magnitude with no. correspond- ing change in field orientation. As will be shown in paper 5 [Kivelson et al., 1973], this increase apparently resulted in the preferential energization of 90 ø pitch angle particles (i.e., betatron acceleration). Associated with this energization were a number of interesting wave particle interactions, as are discussed in paper 5 and also in paper 8 [Scarf et al., 1973]. The exact interpretation of this stage is made complicated by the rapid changes in field configuration that were occurring at this time. We not e from the trend in inclination shown in Figure 3 that Ogo 5 was very close to the magnetic equator and from Table I was just outside the synchronous orbit of ATS 1. As was reported by Cummings et al. [1968], a rapid increase in field magnitude is the typical signature at ATS 1. Thus it would appear that the initial expansion phase changes in field configuration placed Ogo 5 in a position relative to the causative currents that is typical of ATS 1. CONCLUSIONS rro O o 5 MAGNETIC FIELD OBSERVATIONS On the basis of the preceding observations we can draw a number of conclusions about the substorm behavior of the magnetic field in the near tail. For relatively isolated substorms the beginning of the growth phase in the tail corresponds quite closely with the beginning of a southward component of solar wind field at the magnetopause. During the growth

10 phase there is a slow increase in field magnitude in the lobe of the tail. In the cusp there is a McPHERRON ET AL.: SUBSTORM STUDIES, Acknowledgments. This work was supported under the National Aeronautics and Space Ad- corresponding increase and also the develop- ministration contract NAS One of us (M.P.A.) also received support from ESRO and ment of a taillike field. Slow changes in the from NASA under a National Aeronautics and cross-tail component of the field occur, and, in Space Administration International University addition, the plasma sheet appears to thin. Fellowship. This thin plasma sheet can exist at 8 R near ] EI ERENCES midnight. Atkinson, G., A theory of polar substorms, J The onset of the expansion phase is seen in Geophys. Res., 71, , the cusp as the beginning of a decrease in field Aubry, M.P., and R. L. McPherron, Magnetomagnitude and a rotation toward a more di- tail changes in relation to the solar wind magpolar field. Also at this time the plasma sheet netic field and magnetospheric substorms, J. Geophys. Res., 76, , " begins to expand, the lobe field magnitude Aubry, M.P., C. T. Russell, and M. G. Kivelbegins to decrease. Finally, there are sudden son, Inward motion of the magnetopause be- changes in the cross-tail field. fore a substorm, J. Geophys. Res., 75, During the expansion phase the lobe field 7031, magnitude decreases slowly until the boundary Aubry, M.P., M. G. Kivelson, and C. T. Rus- sell, Motion and structure of the magnetoof the expanding plasma sheet arrives, at which pause, J. Geophys. Res., 75, , time there is a sudden decrease. Inside the ex- Aubry, M.P., M. G. Kivelson, R. L. McPher- panding plasma sheethere is a rapid rotation ron, and C. T. Russell, A study of the outer of the field toward a dipole configuration. Asso- magnetosphere near midnight at. quiet and disturbed time, J. Geophys. Res., 76, 5487, ciated with this rotation are large-amplitude Axford, W. I., Magnetic storm effects associated fluctuations in the magnetic field. with the tail of the magnetosphere, Space Sci. DISCUSSION The preceding conclusion represents a considerable generalization of the observations for Rev., 7, , Bame, S. J., J. R. Asbridge, H. E. Felthauser, E. W. Hones, and I. B. Strong, Characteristics of the plasma sheet in the earth's magnetotail, J. Geophys. Res., 72, , only two s.ubstorms. We wish to emphasize, Buck, R., H. West, and R. G. D'Arcy, Satellite studies of magnetospheric substorms on August however, that these conclusions are in agree- 15, 1968, 7, Ogo 5 energetic proton observations, ment with the conclusions derived above from spatial boundaries, J. Geophys. Res., 78, this a review of previou studies. Furthermore, par- issue, ticle observations made on Ogo 5 during these Camidge, F. P., and G. Rostoker, Magnetic field samevents confirm and extend our conclusions. perturbations in the magnetotail associated with polar magnetic substorms, Can. J. Phys., Since our main purpose in this series of papers 48, , is to establish the sequence of events that occurs Coleman, P. J., and R. L. McPherron, Fluctuaduring a single substorm, we have not at- tions in the distant geomagnetic field during tempted a statistical study of the available substorms: ATS 1, in Particles a nd Fields in the Magnetosphere, edited by B. M. McCor- data. We point out, however, thathese con- mac, D. Reidel, Dordrecht, Netherlands, clusions concerning the near tail magnetic field Cummings, W. D., and P. J. Coleman, Jr., Si- are supported by a more inclusive study of a number of substorm events [Aubry et al., multaneous magnetic field observations the earth' surface and at synchronous equatorial 1972]. distance, 1, Bay-associated events, Ra.dio Sci., 3, , At this point we will not attempt a detailed Cummings, W. D., J. N. Ba.rfield, and P. J. description of the substorm sequence for the Coleman, Jr., Magnetospheric substorms obmagnetic field. As. will be shown in subsequent served at the synchronous orbit, J. Geophys. notes, the particle observations on Ogo 5 pro- Res., 73, , Dungey, J. W., The reconnection model of the vide considerable insight into the field behavior. geomagnetic tail, in Earth's Particles and Fields, Con quently, the final paper in this series is edited by B. M. McCormac, Reinhold, New devoted to a synthesis of all the observations, York, Fairfield, D. H., Average magnetic field configuwherein we present the entire substorm se- ration of the outer magnetosphere, J. Geophys. quence and attempt an interpretation. Res., 73, , 1968.

11 3078 McPasaaoN rt A..: SussToa r S ui)ms, 4 Fairfield, D. H., and N. F. Ness, Configuration of the geomagnetic tail during substorms, J. Geophys. Res., 75, , Heppner, J.P., M. Sugiura, T. L. Skillman, B. G. Ledley, and M. Campbell, Ogo A magnetic field observations, J. Geophys. Res., 72, , Hones, E. W., Jr., J. R. Asbridge, and S. J. Bame, Time variations of the magnetotail plasma sheet at 18 R determined from concurrent observations by a pair of Vela satellites, J. Geophys. Res., 75, 4402, Kivelson, M., T. Farley, and M. Aubry, Satellite studies of magnetospheric substorms on Atigust 15, 1968, Ogo 5 energetic electron observations Spatial boundaries and wave particle interactions, J. Geophys. Res., this issue, McPherron, R. L., and P. J. Coleman, Jr., Magnetic fluctuations during magnetospheric substorms, 1, Expansion phase, J. Geophys. Res., 75, , Russell, 12. T., and K. I. Brody, Some remarks on the position and shape of the neutral sheet, J. Geophys. Res., 72, , Russell, C. T., R. L. McPherron, and P. J. Coleman, Jr., Magnetic field variations in the near geomagnetic tail associated with weak substorm activity, J. Geophys. Res., 76, , Scarf, F. L., R. W. Fredricks, C. F. Kennel, and F. V. Coroniti, Satellite studies of magnetospheric substorms on August 15, 1968, 8, Ogo 5 plasma wave observations, J. Geophys. Res., 78, this issue, Sugiura, M., T. L. Skillman, B. G. Ledley, and J.P. Heppner, Propagation of the sudden commencement of July 8, 1966 to the magnetotail, J. Geophys. Res., 73, , Vasyliunas, V. M., Low-energy electrons on. the dayside of the magnetosphere, J. Geophys. Res., 73, , West, H. I., Jr., R. M. Buck, and J. R. Walton, Satellite studies of magnetospheric substorms on August 15, 1968, 6, Ogo 5 energetic electron observations--pitch-angle distributions in the nighttime magnetosphere, J. Geophys. Res., 78, this issue, (Received November 9, 1971; accepted January 5, 1973.)