VOL. 78, NO. 16 •[OURNAL OF GEOPHYSICAL RESEARCH •[UNE 1, 1973

SatelliteStudies of MagnetosphericSubstorms on August15, 1968 2. SolarWind and Outer

R. L. MCPHE'RRON,• G. K. P'ARIiS,•' D. S. COLBURN,3 AND M.D. MONTGOMERY4

We continue the study of the sequer•e of events o.ccurringin the magnetosphereduring several substormson August 15, 1968. We show that the onsets of expansion phas'esidentified in the preceding paper at 0220, 0430, and 0714 UT were each preceded by almost an hour of southward solar wind magnetic field. During the entire interval containing these sub- storms there was no significant change in the solar wind velocity. Intermittent observations of the solar wind particle density and temperature suggestthere were no large changesin the dynamic and static pressure. The fact that the solar wind field remained southward after the onsets of two substorm expansions is interpreted as evidence that the expansion is a process internal to the magnetosphere. However, coincidence of an expansion onset with a large fluctuation of the solar wind field makes it impossible to rule out the possibility that the expansion can be triggered externally. Magnetic field observations in the premidnight sector of the synchronousequatorial orbit show that the 's field here begins to decrease in responseto the beginning of the southwardsolar wind field. Throughout the interval prior to the onset of the expansion (growth phase) the field continues to decrease.During one of the substorms, energetic electron precipitation was observed during this growth phase. In the expansion phase the field at synchronousorbit recovers.

In the first note of this series [McPherro•, expansionswere precededby intervals of south- 1973] we discussedthe importanceof establish- ward solar wind magnetic field and nearly con- ing the sequenceof magnetic changesthat stant solar wind plasma pressure.During both occurswithin the magnetosphereduring a sub- preliminary intervals the magnetic field mag- storm. We noted that it was possibleto estab- nitude at synchronous orbit in the evening lish this sequenceon the basis of existing sector became severely depressed.During one studies, but such a sequencewas subject to substorm for which balloon observations were considerableerror, owing to inconsistenciesin available this depressionwas accompaniedby timing. Consequently,in this set of coordinated energeticelectron precipitation at the conjugate notes we are attempting to establish the sub- point of the synchronoussatellite. storm sequence for two worldwide substorms The general associationbetween southward on August 15, 1968. solar wind field and geomagnetic activity is This second note examines changesin the well established(see review by Hirshberg and solar wind and the outer magnetosphere(syn- Colburn [1969]). Recently, Arnoldy [1971] has chronous orbit, 6.6 Rs) associatedwith these examined this relation statistically using the substorms. Our results show that both substorm hourly auroral electrojet index (AE) and the integral of southwardsolar wind magneticfield. He findsthe parameterbest correlatedwith AE • Department of Planetaw and Space Science is the integral of the GSM southwardZ compo- and Institute of Geophysicsand Planetaw Phys- nent duringthe hour precedingthe maximumof ics, University of California, Los Angeles, Cali- fornia 90024. AE. A similar result has been obtained by Fos- •Geophysics Program, University of Washing- ter et al. [1971], using superposedepoch ton, Seattle, Washington 91805. analysis.Their work showsthat the beginning 8Space SciencesDivision, Ames Research Cen- of a rapid increasein AE is precededby more ter, Moffett Field, California 94035. than an hour of southwardsolar wind magnetic 4Los Alamos Scientific Laboratory, University of California, Los Alamos, New Mexico. 87544. field. Aubry and McPherron [1971] have examined Copyright ¸ 1973 by the American Geophysical Union. the magnetotail responseto both the solar

3054 MCPHERRONET AL.: SUBSTORMSTUDIES, 2 3055 wind magnetic field and magnetosphericsub- events in the solar wind and outer magneto- storms. Using individual substorms,they find spherehave been made by severalinvestigators that the onsetof southwardcomponent of the using different data. They do, however, sug- solar wind magneticfield is followedby an in- gest a sequenceof events which should be creasein the tail field magnitude, a thinning observableduring a single substorm.Our ob- of the plasma sheet, a depressedfield at syn- servations during the two major substorms chronousorbit, and a taillike field on the dawn on August 15, 1968,are in agreementwith these meridian. The start times of these various reports.and show that such a sequenceexists. phenomenawere consistentwith the convection These observationsare discussedbelow. of the southward component past the earth inthe solar wind. Hourly averages andstandard SOLARWIND deviations of various plasma parameters in Observationsof the solar wind from 0000 to the solar wind showedno significantvariations 0900 UT, August 15, 1968, are plotted in Figure correlatedwith the substorms. 1. Plasma parameters shown at the bottom of Cummings et al. [1968] and Coleman and the figure were measuredwith an electrostatic McPherron [1970] have studied the substorm analyzer tm 4A [Hundhausenet al., 1967; magnetic signature at the synchronousorbit. Montgomery et al., 1970]. The approximate The most systematic changesoccur in the H location of Vela during this interval was solar component (parallel to the earth's rotation magnetosphericlatitude --11.6 ø to --3 ø and axis). Near the midnight meridian the field longitude--11.5 ø to 27ø, at a distanceof 18.6 becomesdepressed prior to the onsetof a R•. This location was closeto the noseof the magnetic substorm and recoversin conjunc- magnetosphereand probablywell ahead of the tion with its expansionphase. At the dusk bow shock. meridian the behavior is more complex, de- Magnetic field observationsdisplayed under pending on the individual substorm and the top panel of Figure I were obtainedwith whether it is part of a sequenceof substorms. a fluxgate magnetometeron [Col- If there is any effect,it consistsof a depression burn, 1969]. The approximate location of and recovery in H. The onset of the recovery plorer during these observationswas at the is generally delayed relative to the onset of moon, slightly behind the dawn meridian. A1- the expansionphase at midnight.The beginning though this location is well outside the bow of the depression,however, may occur at any shock,the locationis not well suited for timing time relative to the onset of the expansion the onsets of various substorm phenomenain phase. relation to the solar wind magnetic field. As Enhancedparticle precipitationprior to the was discussedby Aubry and McPherron [1971, onset of a substorm expansionhas been re- appendix 2], there can be considerableam- ported by Hones et al. [1971]. From riometer biguity in the calculatedarrival times of the data these authors have shown that, for at effectsof a solar wind changeat different points least 1 hour prior to the onset of a substorm in the magnetosphere.If we assumethat the expansion,energetic electrons (E • 10-20 key) discontinuity between northward and south- were precipitating throughout most of the ward solarwind magneticfield is perpendicular nightsideauroral oval. This weak activity was to a radius vector, this discontinuity should associatedwith a southwardcomponent of the reach the magnetopauseabout 2 min before solar wind magnetic field, a thinning of the it reachesExplorer 35. If we assumethat the plasmasheet, depressions in H at midlatitudes, discontinuity is perpendicularto the ecliptic and weak bay activity. Pytte and Tre[all and along the Parker spiral, the discontinuity [1972] have also reported electron precipita- would arrive at the magnetopause 15 min tion for more than an hour before the onset before it would arrive at the moon. of negative bays. Using balloon X ray data, Also shown in Figure 1 in the top panel is they show that these events are generally the total field observed by the Ogo 5 mag- weaker and more smoothly varying than the netometerinbound, down from the north lobe largeimpulsive fluxes associated with bay onset. of the tail, on the midnight meridian. A de- The precedingreports of substorm-associatedtailed descriptionof these data is given in 3056 MCPHERRONET AL.' SUBSTORMSTUDIES, 2

io

! ,

4 6 7 8 9 UNIVERSAL TIME

Fig. 1. Solar wind plasma parameters and magnetic field measurements for 0430 and 0714 UT on August 15, 1968, for Vela 66-38 located near the noon meridian and Explorer 35 near the dawn meridian. The vertical dashed lines are the onsets of substormexpansion. The shading on the solar magnetosphericZ component indicates the southward field. The Ogo 5 field magnitudeat •10 RE on midnight meridian is also shownfor comparison. paper 4 of this series.Vertical dashedlines characteristicof this graph is the nearly con- at 0220, 0430, and 0714 UT are the onsets stant solar wind velocity during the entire of substormexpansions as determinedin paper interval. A comparisonwith the onsetsof sub- 1. Vertical arrows at 0330 and 0640 UT are storm expansionsshown by vertical dashed the possible start times for two of the sub- lines reveals no correlation. storm growth phases. The static and dynamicpressures of the solar Plasmaparameters. The solar wind velocity wind for three intervals when the satellite from 0000 to 0900 UT is plotted near the was in the real time data acquisitionmode bottom of Figure 1. The most outstanding are shownin the bottomtwo panels.Note that McPHERRO• E? A•..' SVSSTOR•STvvISS, 2 3057 the static pressurehas been multiplied by and McPherron [1971] for •veral other sub- 10•ø and the dynamic pressureby 108 (units storms. Ho.nes e't a/. [1971] have also con- are dynes/cm2). cludedthat the expansiononset is of internal Although the real time data were not ac- origin from a carefulanalysis of the substorm quired at the most opportunetimes, it is still sequencewith particle data. However, the 0714 possibleto concludethat no large changesoc- UT expansionsuggests the possibilitythat the curred in either pressurebetween the 0430 and expansioncan be triggeredby a suddenchange the.0714 UT expansionphase onsets. in the solar wind field. The expansiononset The observedconstancy of the magnitude occurs almost coincidentwith a very rapid of the solar wind magneticfield and the solar fluctuationin the solar wind field. Whether this wind velocity betweenthe real time Vel• data is coincidentalor indicative of a causal rel•- segmentsand alsothe constancyof the dynamic tion cannot be decided becauseof possible and thermal pressureduring these segments errors in timing. Regardless,the solar wind suggeststhat the solar wind pressuresdid not field remainedsouthward for at least 10 min change significantlyduring the various sub- after the expansiononset, as it did in the storms. These data therefore suggestthat (he precedingsubstorm. substormswere not associatedwith changesin The comparisonof the times identified as the solar wind plasm• parameters. the start of substormgrowth phaseswith the Magnetic field. The three Cartesian corn- southwardcomponent may also be significant. ponentsand magnitudeof the solarwind mag- Vertical arrows at 0330 and 0640 UT are the netic field are shownnear the top of Figure 1. start times determinedin paper 1. Thesetimes The coordinate system used is solar mag- correspondwithin _--+-10min to. the times at nerospheric,the X axis being toward the sun which the solar wind field turned southward. and the earth's dipole axis in the X-Z plane. Consideringthe errors in the determinationof The solar wind field magnitude reflectsthe these times as well as the possibledelays in behaviorof the solar wind velocity,remaining the timing of stagnationpoint effects of • extremely constant thoughout the entire in- parameter measuredat the dawn meridian, the terval. The field components,however, were agreementis quite good. highly variable. Short intervalswhen the field In addition to the variation in the Z corn- pointed southwardare shownby shading.The portent the X and Y componentsalso under- three onsetsof substormexpansions at 0220, went significant changes.In general, the X 0430 and 0714 UT are each preceded by an componentwas. negative throughout the entire hour or more of southward field. interval, and thus • field outward from the A detailedcomparison of the expansiononsets sun was indicated. There was no significant with the behavior of the southwardcomponent changein this componentduring the first two revealsinteresting differences between the three expansiononsets. The Y componentwas posi- events. The 0220 UT expansionwas preceded tire most of the time, as was expected for by several hours of intermittent southward a typical spiral solar wind field line. This com- field. It should be recalled from paper 1 that portent also did not changesignificantly at the this expansiondid not have a clear worldwide time of the first two expansions.It should be signature.In contrast, the 0430 UT expansion noted, however, that both X •nd Y fluctuated was precededby an hour of continuouslysouth- suddenly at the onset of the 0714 UT ex- ward field, and as was discussedin paper 1, pansion. It is possiblethat the simultaneous this expansionwas much more evident in the changein all three componentsmay be related ground magnetograms. to the coincident beginning of the 0714 UT We note also that the 0430 UT expansion expansion. was followedby at least 20 min of southward To summarizethe precedingobservations, we field. This fact suggeststhat the expansionis note the following' the start of the sub- a processinternal to the magnetosphereand storm growth phase coincideswith the begin- not dependenton the solar wind, as was origi- ning of a southward componentof the solar nally suggestedby Heppner eta/. [1967]. wind magnetic field. Substormexpansions are Similar observationswere reported by Aubry precededby • finite interval of southwardsolar 3058 MCPHElmONET An.' SUBSTORMSTUDIES, 2 wind magneticfield. The onsetof the expansion HOUR UT o appearsto be a processinternal to the mag- I netosphere,but at times it may possiblybe triggered by a sudden change in the solar 3 wind field. The magnitudeand the solar mag- netosphericequatorial plane componentsof the field do not play an obvious role in the sub- storm process.

OUTER MAGNETOSPttERE Magnetic field at synchrono•usorbit. Mag- netic field observationsin the outer mag- netosphere are plotted in Figures 2 and 3. 9 IO The satellite ATS 1 on which these measure- II ments were made is located at L -- 6.6 at 12 NO DATA 13

150øW, near the intersection of the geomag- 14 netic and geographicequatorial planes. Local 15 ß midnight is at 1000 UT, whereas dawn, noon, and dusk are 1600, 2200, and 0400 UT, re- 0 I0 20 30 40 50 60 spectively. The V-D-H coordinate system has TIME (MINUTES) its H axis parallel to the earth's rotation axis Fig. 3. Hourly segments of high time reso- andV radiallyoutward in theequatorial plane. lutiondata (15-sec averages) atATS I onAugust 15, 1968. The heavy black dots are expansion The UCLA fluxgatemagnetometer on these onsets. has been described in previous re- ports [Cummingsand Coleman,1968]. The data displayed in Figure 2 are 6-min netic field on August 15, 1968. Dashed hori- averagesof the three componentsof the mag- zontal lines are the componentsof the dipole field at ATS 1. A dashed curve above the H

DAY 228 DATE 08 15 68 component approximates the diurnal variation __l I I I [ I I I I I I I I I I I [ l__ in H observed on quiet days. Vertical arrows D above the H componentare times identified .... as the onset of substormexpansions in paper 1. V¾" The characteristic H component signature -- 2036 of substormsat the synchronousorbit is a de- - pression and recovery in H, as was reviewed - 04300910 I • above. This depressionis apparent for several substormson August 15, 1967, that occurred while ATS i was between dusk (0400 UT) •.:. -" ,o•and midnight (1000 UT). The magnitude of their depressioncan be quite large, e.g., 50 7

_ in the 0430 UT substorm. This depression shouldbe comparedto the _•100-¾magnitude of the quiet time field near dusk.

- I I I I I I I I I I • I I I I I I A comparison of the time variations in H O0 04 08 12 16 20 24 with the onset of the substorm expansions U.T. (HOURS) reveals some of the complexitiesreviewed Fig. 2. Six-minute averagesof the magnetic earlier. The 0220 UT expansionwas followed fieldcomponents at the synchronous satellite ATS almostimmediately by the beginningof a weak 1 onAugust 15, 1968 (V radial,H parallelto depressionin H. The0430 UT expansion,in dipole). The dashed horizontal lines are compo- nentsof dipole field, the dashed curve is the ap- contrast,was preceded byover an hour during proximatequiet day variationin H, and the which the field becamemore and more de- vertical arrowsare expansiononsets. pressedat the satellite. In this case the be- McP•EaaO• ET AL.' SUBSTORMSTUDIES, 2 3059 ginning of the recovery in H at the satellite secondand third substormsdetermined in paper correspondedclosely to the onset of the ex- i started at approximately0330 and 0640 UT. pansion. The 0714 UT event is very similar An examination of the high time resolution to this, except that the duration of the de- ATS i data of Figure 3 showsthat the begin- pressionwas shorter and the recovery of H ningsof the depressionsin H correspondalmost was not complete. exactly to thesetimes. Although this collectionof papers does not A significant feature of the 0640-0714 UT present a detailed study of substormsafter growth phase is more apparent in Figure 2. 0714 UT on August 15, 1968, several im- The beginningof this phasedid not occuruntil portant observationsshould be mentioned for more than an hour after the end of the re- these events. As is apparent in Figure 2, the covery in H associatedwith the precedingsub- partial recovery at 0714 UT was followed by storm. Consequently,there appearsto be little a secondpartial recovery at 0810 UT and a chancethat this depressionis a delayed effect complete recovery at 0910 UT. For these two of this earlier substorm. events, and also for an expansionat 1420 UT, These magneticfield observationsat the syn- there were expansion-associatedvariations in chronous orbit can be summarized as follows. the D component.These variations appear as At about the time identified as the start of sharp, negative 'D spikes' at the top of the the growth phase the H componentbegan to figure. Similar variations in D have been re- decrease. Near dusk this decrease was. tur- ported elsewhere for substorms observed at bulent, whereas near midnight it was quiet. ATS i [Coleman and McPherron, 1970]. At about the time of the onsetof the expansion High-resolution data for the H component pha• the H componentbegan to recover.Near at ATS i are shown in Figure 3. Hourly seg- midnight this recovery was particularly tur- ments plotted from 15-sec averages are dis- bulent. When the satellite was located in the placedsequentially down the page.The vertical late afternoon,a weak expansioncaused a de- scale is shown by an arrow near the top of layed depressionin H. Sharp negative spikes the figure. Heavy black dots on various traces were observedin the D componentfor several representthe onsetof substormexpansions. expansionswhile the satellite was near local These data show additional features of the midnight. correlation of the ATS i field with substorm Ground observationso• electron precipitation expansions.For example, it can be seen that at A TS I conjugate point. Precipitation of the effectsat ATS 1 of the 0220 UT expansion energeticelectrons was measuredat Tungsten, are delayed by as much as 10 min. Similarly, Northwest Territories, Canada (ATS 1 con- the 0430 and 0714 UT recoveriesin H are jugate point, geographic coordinates61ø58'N, delayedby about 3-5 min. 128ø12•W), during the 0714 UT substormex- Another characteristicfeature of substorms pansion of August 15. These measurements at ATS 1 is apparent in these data. In the were made with X ray detectors carried by late evening hours the recovery in H is quite balloonsat approximately3-g/cm 2 atmospheric turbulent (e.g., .0714+, 0810+, and 0910+ depth. Pulse height analysiswas done in two UT). In contrast, the depressionin H is rela- channelshr > 50 kev and hr >_ 20 key. With tively quiet (0640-0714 UT). On the other the assumption of an exponential electron hand, near dusk both the depressionand the energyspectrum the extrapolationof the X ray recovery may be turbulent, e.g., 0330-0430 data throughthe atmosphereusing thick target and 0430-0530UT. More detailedobservations bremsstrahlungtheory givesan averagee fold- of the turbulence near midnight have been ing energy of about 22 kev. Thus for this reported by McPherron and Coleman [1970]. event the electronenergies in the balloon en- Thus far we have discussedonly the cor- ergy channelsaveraged over the spectrumwere relation of the ATS 1 data with the onsetof approximately45 and 75 kev. the expansionphase. However, the fact that One-minute averagesof the count rate for the depressionin H precedesthis onset sug- X rays of >_20 kev are shown in Figure 4. geststhat there is a growth phase at ATS 1 Detectable precipitationabove the instrument as well as at the ground.These phasesfor the thresholdwas first observedat approximately 3060 McPHERRONET AL.' SUBSTORMSTUmES, 2

0700 0730 0800 0850 UNIVERSAL TIME Fig. 4. One-minuteaverages of the countrate of bremsstrahlungX ra.ys (h• •_ 20 kev) measuredwith a balloon-borneX ray detectorat the conjugatepoint of ATS I at Tungsten, NorthwestTerritories, Canada. The contributionof the cosmicray backgroundis about 100 counts/sec.The arrowat 0714UT indicatesthe onsetof the substormexpansion phase.

0653 UT, which coincideswith the growth (MLT) of these observationsaccording to phasededuced magnetically. The X ray fluxes Montbriand[1970] was 2100. increasedsmoothly, reaching a maximum at 0703UT, anddecreased to a minimumat 0713 Co•cnuszo•s UT. An intenseburst of precipitationsud- The observationsdescribed above provide denly appearedat 0715 UT. This increaseis additionalinformation about the two major coincidentwith the onsetof the expansion substormsof August15, 1968.First, they pro- phase. Then the X ray intensity varied vide confirmationof the timeschosen in paper erratically,the flux reachingcosmic ray back- i as the onsetsof the expansionphases. The groundagain at 0812 UT. Anotherweak burst onsetsof recoveriesin the H componentat of precipitationfollowed, reaching a maximum ATS i correspondedwithin severalminutes to at 0816 UT. the chosentimes. Further, for the 0714 UT Visual observationsof were made expansion,visual observationsof the aurora during the X ray measurements.The first at the conjugatepoint of ATS i confirmthat obviouschanges occurred at 0716 UT, when breakupat 2100 MLT occurredat 0716 UT. auroral arcs began to spread and became The visualobservations are supportedby the diffuse.At 0720 UT a westwardsurge passed simultaneousobservations of a suddenincrease overhead.Between 0733 and 0750 UT, auroral in the flux of precipitatedelectrons of energy

activation was to the north. A second interval >20__ kev. of weak activationoccurred between 0814 and The aboveobservations, which showed pre- 0820UT. cipitationto occurprior to the expansionphase, In summary,on August 15, 1968, intense suggestthat trappedparticles in the magneto- precipitationof energeticelectrons was ob- spheremay be redistributedby perturbations servedduring the growth phaseof the 0714 well beforethe auroralexpansion, thus giving UT substorm.Precipitated fluxes increased and further supportof the existenceof the mag- decreasedsmoothly, approximately 12 min netosphericsubstorm growth phase. Such prior to the onsetof the expansionphase. A 'growthphase' redistribution of trappedpar- suddenincrease in this flux, as well as the be- ticles has been noted for events in the dusk ginningof visualauroral activity, occurred at sectorand has been reportedpreviously by the time identifiedin paper i as the onsetof Parks [1970] and Lezniak and Winckler the expansionphase. The magneticlocal time [1970]. McPHERRO• E? AI•.: SVSS?ORMS?vvISS, 2 3061 These observations also suggest that sub- Coleman, P. J., Jr., and R. L. McPherron, storms and the interplanetary southward com- Fluctuations in the distant geomagnetic field during substorms: ATS 1, Particles and Fields ponent of the solar wind field are strongly in the Magnetosphere, edited by B. M. Mc- related and that substorms can apparently Cormac, D. Reidel, Dordrecht, Netherlands, occur when there are no changesin the solar 1970. wind plasma parameters. The observations Cummings, W. D., and P. J. Coleman, Jr., Simul- further indicate that the beginning of an iso- taneous magnetic field variations at the earth's surface and at synchronous,equatorial distance, lated growth phase occurs at about the time 1, Bay-associated events, Radio $ci., 3, 758-761, the solar wind field first turns southward. 1968. Finally, the observationssuggest that the onset Cummings, W. D., J. N. Barfield, and P. J. Cole- of the expansion phase is a process internal man, Jr., Magnetospheric substormsobserved at the synchronous orbit, J. Geophys. Res., 73, to the magnetospherebut that it may at times 6687-6698, 1968. be triggered by a sudden change in the solar Foster, J. C., D. H. Fairfield, K. W. Ogilvie, and wind magnetic field. T. J. Rosenberg, Relationship of interplanetary Since our purposein this collectionof papers parameters and occurrence of magnetospheric is to establishthe sequenceof eventsin a single substorms,J. Geophys. Res., 76, 6971-6975, 1971. Heppner, J.P., M. Sugiura, T. L. Skillman, B. G. substorm, we will not discuss the general Ledley, and M. Campbell, Ogo A magnetic field validity of these observations.However, as was observations, J. Geophys. Res., 72, 5417-5471, noted in the introduction,previous investigators 1967. have noted all these phenomena in separate Hirshberg, J., and D. S. Colburn, Interplanetary field and geomagnetic variations--A unified investigations of different substorms.We do view, Planet. Space Sci., 17, 1183-1206, 1969. point out that there are tremendousvariations Hones, E. W., Jr., S. Singer, L. J. Lanzerotti, betweensubstorms, and consequentlynot all sub- J. D. Pierson, and T. J. Rosenberg, Mag- stormsclearly exhibit the featuresdescribed here. netospheric substorm of August 25-26, 1967, J. Geo79hys.Res., 76, 2977-3009, 1971. Acknowledgments. We wish to acknowledge Hundhausen, A. J., J. R. Asbridge, S. J. Bame, the support of a number of organizations for their H. E. Gilbert, and I. B. Strong, Vela 3 satellite support of various aspects of this work. ATS 1 observationsof solar wind ions: A preliminary data acquisition and reduction were supported report, J. Geophys. Res., 72, 87-100, 1967. by the Office of Naval Research contract N00014- Lezniak, T. W., and J. R. Winckler, Experimental 69-A4)200-4016,ATS I data correction and plot- study of magnetospheric motions and the ac- ting by NASA grant NGL 05-007-004, and the celeration of energetic electrons during sub- time of one of us (R.L.M.) by NASA grant NGR storms, J. Geophys. Res., 75, 7075-7098, 1970. 054)07-305..The balloon X ray measurements were McPherron, R. L., Satellite studies of magneto- supported by NSF GA-487, and one of us spheric substorms on August 15, 1968, 1, State (G.K.P.) thanks Professor J. R. Winckler for of the magnetosphere,J. Geophys. Res., 78, this support of this work while it was carried out in issue, 1973. Minnesota. The principal investigator for the Vela McPherron, R. L., and P. J. Coleman, Jr., Mag- 4 plasma experiment was S. J. Bame of the Los netic fluctuations during magnetospheric sub- Alamos Scientific Laboratory. The Vela observa- storms, 1, Expansion phase, J. Geophys. Res., tions were obtained as part of the Vela Nuclear 75, 3927-3931, 1970. Test Detection Satellite Program, which is jointly Montbriand, L. E., A simple method for cal- sponsored by the Advanced Research Projects culating the local time of corrected geomagnetic Agency of the Department of Defense and the midnight, J. Geophys. Res., 75, 5634-5636, 1970. U.S. Atomic Energy Commission. Montgomery, M.D., J. R. Asbridge, and S. J. Bame, Vela 4 plasma observations near the I•EFERENCES earth's bow shock, J. Geophys. Res., 75, 1217- Arnoldy, R. L., Signature in the interplanetary 1231, 1970. medium for substorms, J. Geophys. Res., 76, Parks, G. K., Acceleration and precipitation of 5189-5201, 1971. Van Allen outer-zone energetic electrons, J. Aubry, M.P., and R. L. McPherron, Magneto- Geophys. Res., 75, 3802-3816, 1970. taft changes in relation to the solar wind mag- Pytte, F., and H. Trefall, Auroral-zone electron netic field and magnetospheric substorms, J. precipitation events observed before and at the Geophys. Res., 76, 4381-4401, 1971. onset of negative magnetic bays, J. Atmos. Colburn, D. S., Description of Ames magnetom- Terr. Phys., 34, 315-337, 1972. eter data from Explorer 33 and 35 deposited in the data bank, report, NASA/Ames Res. Cen- (Received July 9, 1971; ter, Moffett Field, Calif., 1969. accepted January 5, 1973.)