On the Relationship of the Plasmapause to the Equatorward

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 87, NO. All, PAGES 9059-9069, NOVEMBER l, 1982

On the Relationship of the Plasmapauseto the Equatorward Boundary of the Auroral Oval and to the Inner Edge of the Plasma Sheet

J. L. HORWlTZ,1 W. K. Co• 2 C R BAUGHER,3 C R. CHAPPELL3 L A FRANK,4 T. E. EASTMAN,4 R. R. ANDERSON,4 E.G. SHELLEY,5 AND D. T. YOUNG6

ISEE 1 observations of the plasmapause are compared with simultaneous observations of the electron plasma sheet and also the auroral oval observed in DMSP photographs. Only a limited amount of appropriate data was available for the comparisons: the plasmapause/plasmasheet inner edge comparisonswere restricted to the early and late morning sectors, while there were two satisfactory comparisons of the plasmapauseand the equatorward boundary of the auroral oval in the evening sector. However, these examples indicate that the plasmapauselocation often coincides to within zXL- 0.1-0.2 with both the plasma sheet inner boundary and the field line threading the equatorward boundary of the auroral oval. This co-location of the plasmapauseand the plasma sheet inner edge may be due to shielding of the magnetosphericconvection electric field by an Alfv6n layer located at the inner edge of the plasma sheet as discussedby Jaggi and Wolf (1973) and others.

INTRODUCTION sheet population and inhomogeneous ionospheric electrical The locations of both the plasmapauseand the earthward conductivities together cause a modification of the magneto- edge of the plasma sheet are sensitive indicators of large- spheric electric field distribution in which the earthward scale magnetospheric convection electric fields [Axford, edge of the plasma sheet coincides with an 'Alfv6n layer' 1969]. In steady state, the conceptual definition of the that tends to shield most or all of the externally induced plasmapause is that it is the boundary dividing flux tube convection electric field from penetrating into the inner trajectories that encircle the earth, owing to the dominant magnetosphere. Inside this layer the plasma motion may be influence of the earth's corotation electric field, and outside entirely dominated by corotation. Measurements of electric trajectories dominated by the cross-tail electric field. The fields in the ionosphere indicate high electric fields within earth-encirclingflux tubes are filled by plasma flow from the the region of diffuse particle precipitation, with a decrease to ionosphere, while plasma loss at the magnetopauseis approximately zero typically (but not always) seen at the thought to keep the flux tubes outside the plasmapauseat equatorward boundary of this precipitation region (e.g., low density levels. The plasma sheet inner boundary is Gurnett and Frank, 1973]. Identifying the precipitation re- consideredthe most earthward penetration of plasma con- gions with the feet of plasma sheet field lines, the electric vected in from the nightside magnetotail; this plasma is field change would coincide with the inner edge of the diverted around the earth, by corotation and magnetic field plasma sheet. In this case, the abrupt transition from a gradient and curvature drifts, in its transit toward the convection-dominated region to a corotation-dominated re- dayside magnetopause. gion, at the inner edge of the plasma sheet, would corre- Except for zero-energy particles, the simple dawn-dusk spond to the (steady state) plasmapauselocation as well. uniform convection electric field does not lead to coinci- There are surprisingly few published measurements of dence between the plasma sheet inner edge and the plasma- simultaneoussignatures of the plasma sheet inner edge and pause. However, calculations by many authors [Wolf, 1970; the plasmapause, whose location is generally identified Swift, 1971; Taylor and Perkins, 1971; Vasyliunas, 1970, operationally as that of a sharp plasma density gradient 1972; Jaggi and Wolf, 1973; Mal'tsev, 1974; Wolf, 1974; [Chappell, 1972]. Schield and Frank [1970] consistently Harel and Wolf, 1976; Harel et al., 1979, 1981a, b], in observed a separation of 1-5 Re between the plasmapause modeling magnetosphere-ionospherecoupling effects on and the plasma sheet inner edge near local midnight during electric fields, have shown that the presence of the plasma relatively quiet periods. Frank [1971], using a narrow maximum in --•100eV electron fluxes to locate the plasmapause position, observed a coincidence in the plasmapause and • Departmentof Physics,The Universityof Alabamain Hunts- plasma sheet inner edge locations in the midnight-dawn ville, Huntsville, Alabama 35899. 2 Departmentof Physics,Brandeis University, Waltham, Massa- sector, whereas in the evening sector, an 'electron trough' or chusetts 02154. gap was often seen to separatethe two boundaries. Foster et 3 NASA SpaceSciences Laboratory, Marshall Space Flight Cen- al. [1978], in comparing the field line locations for the ter, Alabama 35812. whistler-observed plasmapause density gradient with the 4 Departmentof Physicsand Astronomy, The Universityof Iowa, ISIS 2 observationsof the equatorward edge of the trapped Iowa City, Iowa 52242. 5 LockheedPalo Alto ResearchLaboratory, Palo Alto, California (peak at 90ø pitch angle) and precipitating plasma sheet 94304. electron fluxes, as well as ionospheric densities and other 6 Los AlamosNational Laboratory,Los Alamos, New Mexico parameters, claimed that the equatorward boundary of the 87545. trapped electron fluxes at both dawn and dusk agreed with the whistler-deducedplasmapause. Linscott and Scourfield This paper is not subjectto U.S. copyright.Published in 1982by the American GeophysicalUnion. [1976], examining measurements made near •1ocat dusk, indicated that one plasmapausecoincided with the equator- Paper number 2A1154. ward boundary of the diffuse aurora to within AL --• 0.25.

9059 9060 HORX,VITZET AL' PLASMAPAUSE,PLASMA SHEET, AND AURORAL OVAL

104 _

/• BASEDON ENHANCED •. J- ) NOISE AT UPPER HYBRID • RESONANCEFREQUENCY 18 ...... I ...... •0 ø

14 • H+ .• 1612 H+ 60ø• _' ' '

g r • / • / 1 Z BAS•ON / :.o6[ • V ]•o• LOWERCUTOFF o/ i 1ø2 .1 ...... J Z BEINGATELECTRON o[ ...... 10ø •, = 39.37ø, R - 3.10R E RAM ANGLE (DEGREES) 0055UT,L'5.05, L. T.-22.87, -180 180

lO 1 JANUARY 31, 1978 OUTBOUND : 0033-0O47 UT • UT, L - 3.•, L. T. = 22.01, R = 2.09 - 2.71 Re - • = •.22 ø, R= 2.• RE GM. LAT. = 36.540 - 39.36 ø -180 0 L.T. = 21:23-22:34 lO0 i I I I 2 3 4 5 L --SHELL Fig. 1. Plot of electron density profiles, observed by the plasma wave experiment, versus L shell for the plasmapausecrossing on the January 31, 1978, outboundpass. Estimation of the plasma density was difficult for this case, and two estimates are shown, based on lower cutoff at the plasma frequency and on noise enhancement at the upper hybrid frequency. Roman numeralsrefer to locationsof insetsthat are plots of selectedangular distributions of 0- to 100-eV ion count rates versus spacecraft spin angle about the ram direction. Times and locations of each angular distribution are start times and start locations (- 3-min accumulation intervals). Two separate distributions were seen during the time interval associatedwith II. Drapery-like curve refers to acute angle between magnetic field line and instrument view direction; N and S refer to ions coming from northern and southern hemispheres, respectively.

The purpose of this report is to examine the relationship the auroral oval. Both examples in this section occurred in between the locations of the plasmapause and the plasma the evening sector. In Figure 1 are plotted profiles of sheet inner edge by comparing these boundaries observed in electron density based on two different methods by the data acquired by the ISEE 1 spacecraft; comparisonsof the plasma wave experiment on January 31, 1978 in the evening locations of the plasmapauseand the equatorward boundary local time sector. The angular distributions of 0- to 100-eV of the auroral oval observed in DMSP photographs are also ion fluxes from the plasma composition experiment are also made. ISEE 1 was launched in October 1977 into a highly shown for selected locations along the profile as marked by elliptical orbit with a 30ø inclination to the equator and a 22.5 roman numerals. The decrease in density between L - 3 and Re apogee. The plasmapauseis determined from thermal (0- 4 locates the plasmapausein this range. At about 0040 UT 100 eV) ion measurements made by the plasma composition (R = 2.48 Re, L - 3.84), a transition is observed in the experiment [Shelley et al., 1978] as well as electron density thermal ion distributions in Figure 1, from a broad umbrella- profiles from the plasma wave experiment [Gurnett et al., like shape peaked around the satellite ram direction to a 1978]. In this paper, the plasmapause location is identified doubly peaked distribution with peaks near the maximum from the thermal ion data as the transition from cold, and minimum pitch angles. The broad umbrella curve is a rammed plasma to warm, anisotropic ion distributions. From signature of cold, approximately isotropic plasma being previous reports [Baugher et al., 1980; Horwitz et al., rammed due principally to the spacecraftorbital motion [cf. 1981b], this transition appears to occur along the density Baugher et al., 1980; Horwitz et al., 1981b]; this plasma we gradient traditionally identified with the plasmapause and identify with the plasmasphere. The doubly peaked (field permits a more precise spatially resolved locator of the aligned) distributions are typically observed outside the plasmapause. It is not certain, however, that this operational plasmasphere [Horwitz et al., 1981b]. As discussedabove, definition results in the same location as the conceptual we take the transition between these two types of thermal definition above associated with the boundary of flux tube ion distributions as locating the plasmapause at approxi- trajectories. The inner edge of the plasma sheet is deter- mately 0040 UT, R • 2.48, L • 3.84. mined as a sharp gradient in -1 keV electron fluxes from A DMSP auroral photograph taken over --•15 min centered energetic electron measurements made with the ISEE 1 at 0031 UT on January 31, 1978, is shown in Figure 2 LEPEDEA [Frank et al., 1978]. together with the foot points for the four ISEE 1 locations specifiedin Figure 1. The locations of these foot points were PLASMAPAUSE/EqUATORwARDAURORAL BOUNDARY obtained with a field-line tracing program, using a quiet time In this section we compare the locations of the plasma- Olson-Pfitzer magnetic field model. The foot point locations pause and field lines threading the equatorward boundary of are referenced to the 100-km altitude level typical for aurora. HORWlTZET AL.' PLASMAPAUSE,PLASMA SHEET, AND AURORALOVAL 9061

:-;:;•*i:;:i!!•*.... •:'• :c•,;"

:.:.:., ......

.:½:'.:.:

.:.'.;.z•.

...... '.h•:;. .::..:;

::-..... '"Z;:.... --? ... .7' ..• ::.'z;" -•. .•..

... -"0

...... *A;Z?::"';•::½?-%-," '¾:' .***...... ,•**•.**...?'-"• ;--.<. '"'"*"2" ..,:.-.... .:?';:•'.:-:..;...:'.; ;,..' ......

....

.•.

Fig. 2. DMSPauroral photograph with superimposed ISEE 1 trackand footpoints of locationsspecified in Figures1 and2 for January31, 1978.Time indicated corresponds to timeof satellitelocation at the centerof the photograph.

Evidently the plasmapausetransition location (just poleward Similar data is shown in Figures 3 and 4 for ISEE 1 of foot point II in Figure 2) correspondswell to the equator- outbound pass on March 3, 1978. The• density profile in ward boundary of the rather diffuseauroral oval seenin the Figure 3 showsa slow decreasein the rangeL = 3-6. The DMSP photographat this time. This would indicatethat the transitionfrom plasmasphericto field-aligneddistributions is inner edge of the precipitatingelectrons of the plasmasheet seenfrom Figure 3 to occur betweenlocations II and III, in correspondsto the location of the plasmapause,though the range 0234-0239 UT, L = 5.16-5.80, R = 2.63-2.84 RE. trappedplasma sheet electrons could occur at lowerL shells. The auroral pattern in Figure 4 showsthe equatorward-most 9062 HORWITZ ET AL.: PLASMAPAUSE,PLASMA SHEET, AN•) AURORAL OVAL

Plate I. ISEE 1 LEPEDEA spectrogramfor November 27, 1977. The top four panels display ion counts for the head viewing perpendicular to the spacecraft spin axis, averaged in four sectors referenced to the sun direction. The bottom panel shows full spin-averagedelectron data. HORWITZ ET AL.' PLASMAPAUSE,PLASMA SHEET, AND AURORALOVAL 9063

Plate 2. LEPEDEA spectrogramfor December 4, 1977. 9064 HORWITZ ET AL.' PLASMAPAUSE, PLASMA SHEET, AND AURORAL OVAL

104 MARCH 3, 1978 OUTBOUND 0215 -- 0240 U T R = 1.71 - 2.89 RE GM. LAT. = 34.71 ø - 44.82 ø L.T. = 18:58 - 20:37

103

90ø • = 44.69ø, R = 2.84R E H+ 80 0 180

60o•Z

H+ 102

3oø•

• = 41.10ø, R = 2.27R E •)o - 180 - 120 --60 0 60 120 180 H+ RAM ANGLE 45.36ø, R = 3.05 R lO1 180

•, = 43.70ø, R = 2.63R E

lOo , I [-180 0 180 1 2 3 4 5 6 7

L-SHELL Fig. 3. Similarto Figure 1 for March 3, 1978,outbound pass. feature to be an auroral arc whose brightestpart is to the left equator(rather than h ---38.6 ø) on the sameL shell,as this is of the satellite track. However, close examination shows a where the storm-time ring current causesthe most severe faint continuationof this arc lying betweenII and III, so that inflationof the geomagneticfield. Similarcomments apply to againthe plasmapausematches fairly well with the equator- the case described in Figures 3 and 4. ward-most auroral feature. Althoughthe quiet-timeOlsen-Pfitzer magnetic field mod- PLASMAPAUSE/PLASMA SHEET INNER EDGE el was employedin the foot point specificationabove, it should be noted that the Kp during the January 31, 1978, In both of the previousexamples, the ISEE 1 LEPEDEA event(Figures 1 and2) was4, whereasfor the March2, 1978, was unfortunately not operating during the plasmapause case(Figures 3 and 4) Kp -• 4+. Inspectionof the magnetic crossings.In this sectionwe examine casesin which the field contourmaps of Sugiuraet al. [1971]near midnight LEPEDEA was on and was able to specify the earthward indicatesa differencebetween magnetic field magnitudesfor edge of the plasma sheet. Unfortunately, for the period Kp = 0-1 and Kp = 2-3 of _<53' for the R = 2.48 Re, X = November 10, 1977,to April 10, 1978,examined here, we did 38.6ø location of the plasmapausecrossing on January 31, not find any casesin which datato obtainthe DMSP auroral 1978. This deviation is less than 0.5% of the total field boundary,plasmapause, and LEPEDEA plasmasheet inner magnitudethere. Also, similarlysmall magnitude deviations edge were all available. are evident in the differences in the field line traces of Mead Figure $, for the outbound ISEE 1 pass in the early andFairfield [ 1975]between Kp = 0, 0+ andKp >- 3 for their morning sector on November 27, 1977, shows the total innermostplotted field lines passing near synchronous orbit; densityprofile, selected0- to 100-eV ion distributions,and for the more earthward locations considered here the devi- electron count rates from the 1.1-keV channel of the 4E ationsare expectedto be even smaller.Though actually the LEPEDEA detector, whose aperture lies in the spacecraft displacementin foot-pointlocation results from inflation and spin plane [Frank et al., 1978].During this traversal,Kp = bendingof field lines, even if the magnitudedeviation is as 2-2-. In this example,the plasmapauseis identifiableas the highas 1-2% for the Kp hereof 4-5, thefoot-point uncertain- steepdensity gradient at L = 4.84, whereagain signatures of ty shouldbe muchsmaller than the distancebetween points the thermal ion distributions exhibited the transition from II and III in Figure2 as well as the uncertaintyin prescribing cold, isotropicplasma to warm, field-alignedions. Simulta- the location of the rather diffuseequatorward auroral bound- neous with the plasmapauseencounter was the observed ary. Thus, we do not expect the use of the Olsen-Pfitzer strongincrease in keV electronfluxes which is identifiedas model to introducegreater uncertainties than are otherwise the inner edge of the electronplasma sheet. As is indicated presentin this semi-quantitativecomparison. It shouldbe by the energy-timespectrogram of Plate l, energydispersion noted, however, that the inaccuracyin the field line tracing was in fact presentin the encounterwith the electronssuch would be more pronouncedif the satellitehad been at the that whereas the sharp rise in electron fluxes occurred HORWITZET AL..' PLASMAPAUSE,PLASMA SHEET, AND AURORALOVAL 9065

ISEE-1 TRAC'K

¸. /

Fig. 4. Similar to Figure 3 for March 3, 1978. essentially simultaneously at --•1220 UT (L --• 4.84) for crease in 1.1-keV electron fluxes, this example of Figure 5 is electron energies ---0.2-2 keV, the enhancement of higher a particularly striking instance of its correspondence with energy electrons appears at later times, --•1230 UT (L --• 5.6) the plasmapause. for the 5- to 8-keV electrons. Similar dispersionis evident in A further comparison of the plasmapause and plasma the data of Schield and Frank [1970]. However, if we do sheet inner edge is contained in Figure 6 for the outbound identify the plasma sheet boundary through the sharp in- ISEE I pass on December 4, 1977, during which Kp = 4. In 9066 HORWITZ ET AL.' PLASMAPAUSE,PLASMA SHEET, AND AURORALOVAL

300 lO4 _

DENSITY

- 1.1 key ELECTRON e.---e

_ COUNT RATE

_ i ',,/ _ i 1223 UT, L = 5.18, L. T. = 3.56, 103 -- )• = 34.00ø, R = 3.54R E ..--.. 90ø I o o

I 60ø 0 I

H+ I z i i o I-. 300m z keV k- 102 _ 1212 UT, L = 4.25, L. T. = 3.17, o z )• = 31.93ø, R= 3.07R E 0ø z o • - -180 -120 --60 0 60 120 180 • - RAM ANGLE (DEGREES) 1218UT, L = 4.72,L. T. = 3.38, 0.8 key • -- 33.10ø, R = 3.31R E

-180 o 180 -lOO >

H+ 101 -- NOVEMBER 27, 1977 _ OUTBOUND _ 1152 - 1230 UT _

- GM.R= 2.27-LAT. =3.81 •5.67 Røe - 34.60 ø

- 1221UToL = 4.99, L.T. = 3.49, L. T. = 02'10 - 03:50 ,,,•,"'•''"• 0.7 key - )• =33.65 ø,R = 3.44 RE •,•"• -18o..... 6 ..... 18oe...e...,e.,..,• '• 10 I , • o I 2 3 5 6

L-SHELL

Fig. 5. Similar to Figures 1 and 4 for the November 27, 1977,ISEE 1 outboundpass, with LEPEDEA 1.1-keV spin planeelectron count rates also displayed. To improveenergetic electron L shell/timeresolution near the plasmasheet crossing(beyond the ---2 minLEPEDEA cycletime for the 1.1keV channel),we haveincluded three additional points at 0.7, 0.8 and 0.9 keV near L = 4.8. this case the plasmapausemay be seen from the density keV electron fluxes in all local time sectors.) The one gradient and ion distribution transition to occur in the range exception was a case in which there was a transition to -1-2 L = 4.07 - 4.27, and examination of higher resolution ev (as opposed to •10 eV) field-aligned flowing ions that thermal ion spin curves places the plasmapausetransition at preceded the plasma sheet inner boundary crossing. This L = 4.14. The keV electron fluxes in this example exhibited caseoccurred during a magneticallyquiet period,and these a rise from L = 3 and theh plateau up to the L --• 4 location of cold flowing ions were probably associatedwith filling of the theplasmapause, followed by a furthersharp increase at the outer plasmasphere [Horwitz et al., 1981b]; in fact, if we had plasmapauselocation. Both flux increasesca n be distin- taken the transition between these 1-2 eV flowing ions to guished in the electron portion of the spectrogramin Plate 2. • 10 eV field-alignedions to be the plasmapause,this would There is thus someambiguity i n the plasmasheet inner edge give coincidencewith the plasmasheet inner edge for this determinationfor this event.However, we selecthere the case as well. In most of the other cases the plasmapause rise at L = 4.14 owing to the sharper rise of --• 100 counts/s transition Coincidedwith a transition to warm (• 10 eV) field- within fiaL--• 0.4; the enhancedelectron fluxes seen in L = 3- aligned ions; in two instances, a transition to pancake 4.13may be long lifetime electrons [Schield and Frank, 1970] distributions (peak fluxes at 90ø pitch angle) in warm ions deposited dbring a prior injection or more earthward move- [Horwitz et al., 1981a] occurred instead. We note that the mentof the plasmasheet. coverage for both boundaries is seen from Figure 7 to be The locations of the plasma sheetinner boundariesand the limited to the midnight-to-noonsector. plasmapause transitions (defined as the outer boundaries of DISCUSSION cold, isotropic plasma) are plotted versus L shell and local time in Figure 7, for the passesduring the November 1977- From the small sampleof comparisonspresented in this April 1978 interval when both locations were obtained. report it is evident that frequently, at least, the plasmapause There were fourteen passesin which both boundaries were (determined from the transition in thermal ion distributions) determined,and in all casesexcept one the boundaries and the inner edge of the electron plasma sheet(and the field coincided to within z•L - 0.1. (The term 'plasma sheet' inner lines threading the equatorward boundary of the aurora) are edge is employed here to designatethe earthward edge of the co-located. This conclusion differs with the results of Schield HORWITZ ET AL.: PLASMAPAUSE,PLASMA SHEET, AND AURORALOVAL 9067

lO3 2OO DENSITY

1.1 keV ELECTRON / COUNT RATE / / 90ø 5.7 KeY

4.1 key

1•07 UT, L ' 2.99, L. T. ' 2.3•, ! •, ' 17.74O, R - Z73R E 2.1 key I,- -180--120 --•0 0 •0 120 180 2.9 key z RAM ANGLE (DEGREES) 1829 UT, L - 4.07, L. T. - 3.0•, / 1.5 keV 100 •, - 21.51ø, R - 3.52R E / z

1629 UT, L - 4.27, L. T. - 3.18, • 21.93ø, R 3.M R 0 180

DECEMBER 4, 1977 OUTBOUND 1558-1635 UT R = 2.44- 3.92 RE GM. LAT. 14.82 ø .-22.58 ø LT = 02:04 - 03:25 . .•,.. •..•o: R, •.7,.. ".E. -180 1•12UT,L'4.45,o L. T.-3.17, 1•o ] I , I , I 4 L-SHELL

Fig. 6. Similar to Figure 5 for December4, 1977,outbound passß

and Frank [ 1970]who reporteda typical separationof 1-5 Re electric field associated with the inner edge of the plasma between the plasma sheet inner edge and the plasmapause. sheet. Whether such an electrical shielding results from Part of this discrepancycould be ascribedto differencesin diversion of gradient drift currentsinto the ionosphereat the the identificationof the plasma sheet location. Schield and inner edge of convecting plasma sheet [Jaggi and Wolf, Frank [1970] used a gradient in electron energy density in 1973] or from ionospheric conductivity gradients at the range90 eV-50 keV, whereaswe have useda sharpgradient equatorward boundary of a ring of enhanced conductivity in keV electron fluxes to identify the 'plasma sheet' inner [Vasyliunas, 1970] is not critical here since in both caseswe edge. In dispersiveevents such as the event in Figure 5 and expect the edge of the shieldingregion to be on field lines Plate 1, the 0.09- to 50-keV energy densitygradient would threading the inner edge of the plasma sheetand the equator- place the plasmasheet edgebeyond the plasmapause.The ward boundary of the auroral oval. An alternate explanation proper choice of identifier dependson the physicsunder for the correspondencewould be that the cold plasmaspheric consideration, but the choice here would seem to yield the plasma limits the earthward penetrationof the plasma sheet earthward-mostpenetration of fresh plasmasheet electrons through wave-particle induced precipitation of energetic and is likely to be appropriate for many considerations. particles [e.g., Thorne et al., 1973]. Another difference is that Schield and Frank [ 1970] included Although we are not aware of any reliable plasma flow a predominanceof quiet periods; the single instance in measurements made at high altitudes across the plasma- Figure 7 where we have indicateda separationwas sucha pause, measurementsof ionosphericelectric fields typically quiet period, althoughthis separationprobably spanned a fall abruptly to small values in equatorwardcrossings of the plasmaspherefilling region. A third differenceis that Schield lower latitude boundary of plasma sheet precipitation [Gur- and Frank [1970] included many casesin the premidnight nett and Frank, 1973; Horwitz et al., 1978], although during sector,whereas the casesreported here were in the early and active periods occasionallyvery high-speedflows and elec- late morning sector. In fact, the report by Frank [1971] tric fields have been observed in the trough region equator- indicated (also using the energy density gradient indicator ward of the auroral precipitationboundary [e.g., Spiro et al., for the plasmasheet edge) that in the post-midnightsector 1979]. In such latter cases, it would be expected that a the plasmasheet edge was coincidentwith the plasmapause, plasma trough or gap region might develop between the which agreeswith the presentresults, while a gap is present plasmapauseand the plasma sheet inner edge, becausethe in the premidnight sector. high-speed convection would erode the plasmapause to As discussedin the introduction, a correspondencebe- lower L values [e.g., Chappell, 1972]. Such a gap might tween the plasmapauseand plasmasheet inner edgemay be appear most commonly in the evening sectoras suggestedby understood in terms of shielding effects on the convection the results of Frank [1971]. Alternately, this evening sector 9068 HORWITZ ET AL.: PLASMAPAUSE,PLASMA SHEET, AND AURORAL OVAL

ß PLASMA SHEET INNER BOUNDARY

OUTER BOUNDAR 12 LT OFPLASMACOLD, ISOTROPIC PANCAKE DISTRIBUTH 8 2 3 4 5 6 7 OUTSIDE L-SHELL COLD PLASMA 00

Fig. 7. Comparisonof plasmapauseand earthward edge of electronplasmasheet locations for ISEE 1 passes.

'gap' might be the plasmaspherefilling region in many cases. such as --• 60 ions/cc [e.g., Maynard and Grebowsky, 1977] As these last commentsindicate, it is important to extend the would appear to be rather ad hoc and may at times give present study to all local time sectorsand also to investigate misleading results. As with the plasma sheet inner edge the effects of magnetic activity upon the relationshipbe- indicator discussedearlier, the choiceof operationalplasma- tween the plasmapauseand the plasmasheet inner edge. An pauselocator dependsupon the physicalphenomenon under opportunity to accomplishthis may comewith data from the consideration, and we suggestthat in many instancesthe Dynamics Explorer spacecraft. locator employed here be consideredby other researchers As a final note, we reiterate that we have used as our working on plasmapause-referencedphenomena. plasmapauseindicator a method not generallyused, i.e., the observed transition in the dominant thermal ion distributions Acknowledgments. This research was supported in part by NASA contracts NAS8-30563 and NAS8-33982 at the University of from cold, quasi-isotropicdistributions to warm, anisotropic Alabama in Huntsville, by NASA contract NAS5-20538 at Lock- distributions(usually field aligned). If desired, this method heed, by Swiss National ScienceFoundation grant 2.886.77 at Bern, can often yield a plasmapauselocation with a time resolution and NASA contracts NAS5-20093 and NAS5-20094 at the Universi- of the order of the satellite spin period, which is 3 s in the ty of Iowa. case of ISEE 1, and translates(for typical satellitevelocities The Editor thanks R. A. Heelis and J. C. Foster for their assistancein evaluating this paper. near the plasmapause)to a resolutionof •< 15 km or A L •< 0.003 (although we have soughtonly AL --• 0.1 resolution in REFERENCES this report). Although positive spacecraft potentials may Axford, W. I., Magnetosphericconvection, Rev. Geophys. Space mask cold ion distributions, particularly in low plasma Phys., 7, 421, 1969. density regions beyond the plasmapause,this indicator of- Baugher, C. R., C. R. Chappell,J. L. Horwitz, E.G. Shelley, and fers enhancedprecision relative to, for example,the designa- D. T. Young, Initial thermal plasmaobservations from ISEE 1, tion by the sharp gradient method. We estimate that this Geophys. Res. Lett., 7, 657, 1980. latter method would yield L shell resolutions of approxi- Chappell, C. R., Recent satellitemeasurements of the morphology and dynamicsof the plasmasphere,Rev. Geophys.Space Phys., mately AL - 1, 2, 0.1 and 0.1, respectivelyfor the January 10, 951, 1972. 31, March 3, November 27, and December 4 crossings Foster, J. G., C. G. Park, L. H. Brace, J. R. Burrows, J. H. presented in Figures 1, 3, 5, and 6. The present indicator Hoffman, E. F. Maier, and J. H. Whitaker, Plasmapausesigna- may be modified so as to include within the plasmapause tures in the ionosphereamd magnetosphere,J. Geophys.Res., 83, 1175, 1978. occasionally seen regions of 1- to 2-eV flowing ions most Frank, L. A., Relationship of the plasma sheet, ring current, probably identified with the plasmaspherefilling, as men- trapping boundary and plasmapausenear the magneticequator tioned above. Indicators based on particular density levels and local midnight, J. Geophys. Res., 76, 2265, 1971. HORWITZ ET AL.: PLASMAPAUSE,PLASMA SHEET, AND AURORALOVAL 9069

Frank, L. A., D. M. Yeager, H. D. Owens, K. L. Ackerman, and M. Maynard, N. C., and J. M. Grebowsky, The plasmapauserevisited, R. English, Quadrispherical lepedeas for ISEE's 1 and 2 plasma J. Geophys. Res., 82, 1591, 1977. measurements, IEEE Trans. Geosci. Electron., GE-16, 221, 1978. Mead, G. D., and D. H. Fairfield, A quantitative magnetospheric Gurnett, D. A., and L. A. Frank, Observed relationshipsbetween model derived from spacecraft magnetometerdata, J. Geophys. electric fields and particle precipitation, J. Geophys. Res., 78, Res., 80, 523, 1975. 145, 1973. Schield, M. A., and L. A. Frank, Electron observations between the Gurnett, D. A., R. R. Anderson, F. L. Scarf, R. W. Fredericks, and inner edge of the plasmasheet and the plasmasphere,J. Geophys. E. J. Smith, Initial results from the ISEE-1 and -2 plasma wave Res., 75, 5401, 1970. investigation, Space Sci. Rev., 23, 103, 1978. Shelley, E.G., R. D. Sharp, R. G. Johnson,J. Geiss, P. Eberhardt, Harel, M., and R. A. Wolf, Convection, in Physics of Solar- H. Balsiger, G. Haerendel, and H. Rosenbauer,Plasma composi- Planetary Environments, vol. 2, edited by D. J. Williams, p. 617, tion experiment on ISEE-A, IEEE Trans. Geosci. Electron, GE- AGU, Washington, D.C., 1976. l6, 266, 1978. Harel, M., R. A. Wolf, and P. H. Reiff, Computer modeling of Spiro, R. W., R. A. Heelis, and W. B. Hanson, Rapid sub-auroral events in the inner magnetosphere,in Quantitative Modeling of ion drifts observed by Atmospheric Explorer C, Geophys. Res. the MagnetosphericProcesses, Geophys. Monogr. Ser., vol. 21, Lett., 6, 657, 1979. edited by W. P. Olson, p. 499, AGU, Washington,D.C., 1979. Sugiura, M., B. G. Ledley, T. L. Skillman, and J.P. Heppner, Harel, M., R. A. Wolf, P. H. Reiff, R. W. Spiro, W. J. Burke, F. J. Magnetospheric-field distortions observed by OGO 3 and 5 J. Rich, and M. Smiddy, Quantitative simulationof a magnetospher- Geophys. Res., 76, 7552, 1971. ic substorm, 1, Model logic and overview, J. Geophys. Res., 86, Swift, D. W., Possiblemechanisms for formation of the ring current 2217, 1981a. belt, J. Geophys. Res., 76, 2276, 1971. Harel, M., R. A. Wolf, R. W. Spiro, P. H. Reiff, C.-K. Chen, W. J. Taylor, H. E., and F. W. Perkins, Auroral phenomenadriven by the Burke, F. J. Rich, and M. Smiddy, Quantitative simulation of a magnetosphericplasma, J. Geophys. Res., 76, 272, 1971. magnetosphericsubstorm, 2, Comparison with observations,J. Thorne, R. M., E. J. Smith, R. K. Burton, and R. E. Holzer, Geophys. Res., 86, 2242, 1981b. Plasmaspherichiss, J. Geophys.Res., 78, 1581, 1973. Horwitz, J. L., J. R. Doupnik, and P.M. Banks, Chatanika radar Vasyliunas, V. M., Mathematical models of magnetosphericcon- observations of the latitudinal distributions of auroral zone elec- vection and its couplingto the ionosphere,in Particles and Fields tric fields, conductivities and currents, J. Geophys. Res., 83, in the Magnetosphere, edited by B. McCormac D. Reidel, 1463, 1978. Hingham,Mass., 1970. • ' Horwitz, J. L., C. R. Baugher, C. R. Chappell, E.G. Shelley, and Vasyliunas, V. M., The interrelationshipof magnetosphericpro- D. T. Young, Pancake pitch angle distributions in warm ions cesses, in Earth's Magnetospheric Processes, edited by B. observed with ISEE 1, J. Geophys. Res., 86, 3311, 1981a. McCormac, p. 29, D. Reidel, Hingham, Mass., 1972. Horwitz, J. L., C. R. Baugher, C. R. Chappell, E.G. Shelley, D. T. Wolf, R. A., Effects of ionosphericconductivity on convectiveflow Young, and R. R. Anderson, ISEE 1 observations of thermal of plasma in the magnetosphere,J. Geophys.Res., 75, 4677, 1970. plasmain the vicinity of the of the plasmasphereduring periods of Wolf, R. A., Calculations of magnetosphericelectric fields, in magnetic quieting, J. Geophys. Res., 86, 9989, 1981b. Magnetospheric Physics, edited by B. M. McCormac, p. 167, D. Jaggi, R. K., and R. A. Wolf, Self-consistent calculation of the Reidel, Hingham, Mass., 1974. motion of a sheet of ions through the magnetosphere,J. Geophys. Res., 78, 2852, 1973. Linscott, G. R., and M. W. J. Scourfield, Observations of the plasmapauseand diffuse aurora, Planet. Space Sci., 24, 299, 1976. Mal'tsev, Yu. P., The effect of ionospheric conductivity on the (Received April 19, 1982; convection system in the magnetosphere,Geomagn. Aeron., 4, revised July 12, 1982; 128, 1974. accepted July 27, 1982.)