Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.izmiran.rssi.ru/magnetism/Feldstein_2012/PDF/364.pdf Äàòà èçìåíåíèÿ: Wed Mar 11 16:37:04 2009 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 21:09:45 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: viking 2

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 100, NO. A9, PAGES 17,429-17,442, SEPTEMBER 1, 1995

Structure of auroral precipitation during a theta aurora from
multisatellite observations Y. I. Feldstein, P. T. Newell,2 I. Sandahl, J. Woch,3'4 V. Leontjev, 1 3 S. · and V. G. Vorobjev ·
Abstract. A 0 aurorapreviously discussed the basisof Viking images(northern on hemisphere) DE 1 images and (southern hemisphere) reexamined light of additional is in data,primarilythe auroral plasma distribution determined as from the Viking, DMSP F6, and DMSP F7 satellites.This event, which occurredbefore a substorm expansionphase on August3, 1986, appeared the imagesto consist a singlearc alongthe morning in of sidein the northern hemisphere alongthe eveningsidein the southern and hemisphere and was isolated from the auroraloval in both setsof images.On the basis of the auroral plasma distribution inferredfrom threesatellites, brightest the arcsdo occurat the locations indicatedby the imagers, the arcsare in fact connected the main oval but to with continuous precipitation, weakersecondary (not observed the imagers) and arcs by occurin the opposite hemisphere magnetically conjugate the brightarcs.Theseobserto vationssupport interpretation the 0 auroraas occurring closed the of on field linesas a resultof the expansion the morning of andeveningsector ovalsinto the polar cap.A careful examinationof the characteristics the observedauroral energy plasma suggests of additional conclusions. appears the ionospheric It that manifestation the recently of discovered low-energy electron layercan be identified with a complicated structure of softprecipitation the poleward at edgeof the mainprecipitation region. Finally,unlike recentreports, ionswerenot observed havea cutoffin the polarcap that is any the to sharper than that of the electrons.
1. Introduction

The 0 aurora has been interpreted representing bias a furcationof the plasmasheet[Frank et al., 1986], an interpretation that implies a regionof open field lines separatingthe 0 aurorafrom the main auroraloval. Conversely, Meng [1981], Murphree and Cogget [1981], and Lundin et al. [1991] have interpretedpolar arcs as representing an expansion the closedauroralfield lines poleward,espeof cially in the dawn and dusk sectors. Hones et al. [1989] further developedthis interpretationobservationallyand theoretically and showedthat a "horse-collar" auroralpattern, in which two "bars" or sun-alignedpolar arcs exist alongthe polewardedgeof the expanded morningside and eveningside auroral ovals, is a commonsituationfor magnetically quiet times. Observational support thesemodelshas been mixed. for Mizera et al. [1987], usingprecipitation data from NOAA 7 (southernhemisphere)and DMSP F6 (northernhemiRadiowavePropagation, Troitsk, Russia.

sphere)satellites, found that bursts,which they believedto be 0 auroras,occurredsimultaneously both hemispheres. in Obara et al. [1988] concludedthat a polar arc imaged by Viking in the northern hemispherewas conjugateto particle measurements from Exos C measurements in the

southernhemisphere.These results supportan expansion of closedauroral field lines polewardto explain polar arcs
and 0 aurora. However, Craven et al. [1991] studied an in-

stance of 0 aurora before the expansion phase of a substorm on August 3, 1986, using the Viking (in the northern hemisphere) and DE 1 (in the southernhemisphere) imagers. This occurrenceis the only known example of imagingboth hemispheres duringa 0 aurora.The imagesindicatedthat a single sun-aligned polar arc existed along the morning (evening) sectorin the northern(southern) hemisphere, with both arcs separated from the main auroral oval by regions without luminosity.These results of Craven et al. [1991] support the interpretation Frank of et al. [1986] of the 0 aurora as representing bifurcation a the 1Instituteof Terrestrial Magnetism,Ionosphereand of the plasma sheet with open field lines separating
0 aurora from the auroral oval.

9'Applied Physics Laboratory, Johns Hopkins University,
Laurel, Maryland.

3Swedish Institute Space of Physics, Kiruna, Sweden.
4Now at Max-Planck-Institut far Aeronomie, Lindau,
Germany.

5Polar Geophysical Institute, Apatity, Russia.
Copyright1995 by the AmericanGeophysical Union.
Papernumber95JA00379.
0148-0227/95/95 JA-00379505.00

The distributionof auroral energy plasma (hundredsof eV to several keV) in the magnetosphere among the is mostimportantcharacteristics the globalmagnetospheric of morphology. present,the large-scale At. structure and dynamics of auroral luminosity and particle precipitationin the ionosphere and the issue of mappingthese regionsto the outer magnetosphere under active discussion are [e.g., Stasiewicz, 1991; Weiss et al., 1992; Elphinstone et al., 1993a]. In this paper we reexaminethe August 3, 1986, event using particle data from the DMSP F6 and F7 and Viking satellites investigate to theseand relatedissues. Be-

17,429


17,430

FELDSTEIN ET AL.: THETA AURORA FROM MULTISATELLITE OBSERVATIONS

region, which is poleward of the main auroral oval, is termed "soft precipitation"(SP). In magneticallyquiet intervals,polar arcs may exist up view of current understanding particle of precipitation to very high latitudes, filling up much of the polar cap. redetermining polewardboundary the gions and their relationshipto high-altitudestructures on This factor complicates the nightsideis presented the remainder this section. of the auroral oval and the softer precipitationregion on in of data only, sincetheir spectra The structureof the high-altitudemagnetotail has been the basisof auroralelectron established from comprehensive situ measurements in [e.g. in these two regions above the discreteauroral forms are Eastman et al., 1984; Parks eta!., 1992]. It contains the very similar [Murphreeet al., 1983; Petersonand Shelley, central plasma sheet (CPS), the main body of hot, nearly 1984; Frank et al., 1986]. Arguing for the existenceof differences betweenion precipitation the auin isotropicplasma in the tail; the plasma sheetboundary systematic layer (PSBL) with field-aligned,velocity-dispersed particle roral oval and poleward of it, Troshichevand Nishida that beams at the outer edge of the CPS; and a low-energy [1992] and Gusev and Troshichev[1992] suggested electronlayer (LEL) at the outer edge of the PSBL with the ion precipitation drop-offallowsa betterdetermination an outward electron beam and an earthward beam of ions, of the auroral oval polewardboundary.These authorsarboth at energiesof a few hundredeV or less.The CPS is gued that the ion flux levels dropped off much more located on each side of the neutral sheet in the sharplypolewardof the auroraloval than did the electrons. With development a magneticdisturbance polar of the magnetotail,on magneticfield lines that are stretched in but the antisunwarddirection. The earthwardboundaryof the arcspolewardof the auroraloval disappear, poleward CPS lies close to the boundaryof stabletrappingfor high- of the highestlatitudeauroraloval arc a narrowband of remains.The result is a faint luenergy particles(or the isotropicboundary).On the night- soft particleprecipitation side it defines a narrow transitionshell region betweenthe minosity, which has been called polar diffuse aurora quasidipolar magnetic of theinner field magnetosphere (PDA) [Eather and Akasofu, 1969]. Precipitationin this reand Because the relativelylow of taillike, stretchedfield region farther down the tail. This gion as a rule is structured. the is transitionshell region is due to the inner edge, or a strong intensityof PDA precipitation, luminosity subvisual, outwardgradient,of the integratedcross-tailcurrentof the andthe spectrum softer is thanin the auroral oval'As a result the luminosityis at higher altitudesand is redder. plasmasheet. The region of the inner magnetosphere extendingin- Such luminosity was identified in pioneeringpapersby ward from the trapping boundaryto the zero-energycon- Eather [1969] and Eather and Mende [1972], basedon airobservations with relativelypoor spatial vectionboundary,or the plasmapause, namedthe rem- bornephotometric was Despite the name, there is evidencethat PDA nant layer (RL) by Feldstein and Galperin [1985, 1993] resolution. and Elphinstoneand Hearn [1993]. Continuous convection includes structured fluxes. The character of the auroral luprecipitating soft auroralelectron and plasma injection during substorms transportlow- minosityand structured energy particles into this region. Therefore they can be fluxes poleward of the large-scalepolar arcs were deconsidered remnants the plasma as of sheetwithinthe r·e- scribedby Austin et al. [1993] using Viking data. The width of this region varies from several tens to several gion of energetic particletrapping. The authors have discussed'their view of magneto- hundredsof kilometers. Relying on particle data taken spheric topology and.how these high-altitudestructures from the DMSP satellites,Akasofu et al. [1991] also conexistspoleward map to the ionosphere some.length at elsewhere [Feldstein cludedthat a belt of particleprecipitation
cause of the recent advancements in understanding the

mappingof the magnetotailinto the auroraloval and to make our interpretation the particle data clearer, a reof

and Galperin, 1985; Ga(perin and Feldstein, 199!;
Feldstein eta!,, 1994a], Those who ·wish more:detail on

of the auroral oval. Inside this belt the auroral luminosity

is soweakthatit is usually detectable thepresently not in magnetospheric t.op01ogy and'its mappingto the iono- available satellite auroral images. At PDA latitudes ion with velocity sphereare referred to these papers.., 'Because the com.- fluxes exist with E<10 keV and sometimes of
dispersed ion, structures (VDIS 2) signatures characterized toward lower latitudes[Zelenyiet herein refer to the precipitationstructure neutral(purely by ion energy decreasing ..in descriptive)terms that do not contain implicit mapping as- al., 1990]. The occurrence bright polar arcs in the polewardreof sumptions be sure that our views of the mappingsugto to field topolgestedby the evidence are included but not embeddedin gion is closely connected the geomagnetic the nomenclature. ogy. Observations have shownthat the plasmacausingthe Broadly speaking there exist three major regions of luminosity of large-scalepolar arcs lies on closed field lines exhibiting a double loss cone type of distribution nightsideprecipitation(in order of increasing latitude): 1. This first regionis the diffuseauroralelectron pre- characteristicof a trapped particle population [Peterson cipitation on the background of the outer radiation belt and Shelley, 1984; Frank et al., 1986; Eliasson et al., electrons to the stabletrappingboundary. up Through 1987]. Three possible morphologiesof closed magnetic much or sometimesall of this region the electron energy field lines in the polewardregionhave been suggested: (1) increases with latitude. This region will be termed "diffuse wideningof the plasmasheetand luminosity regiontoward higherlatitudeson the dawn and dusk sides[Meng, 1981; precipitation"(DP). 2. The second region is hard structuredelectron and Murphree and Cogger, 1981]; (2) bifurcationof the open capby a wedge closed of field isotropicion precipitationbetween 1 and 10 keV, poleward field linesof the polar originate[Frank et of the'stabletrapping boundary. This regionwill be termed lines from which the bright emissions al., 1986; Frank and Craven, 1988]; and (3) an expansion "hard precipitation"(HP). 3. The last region is soft structuredelectronprecipita- of the low-latitude boundary layer (LLBL) [Lundin al., et tion with low intensity and typicalenergies 1 keV. This 1991]. In the first and third cases, one might expect an <

plex structurethemagnetOspheric sheet, will of plasma We


FELDSTEIN AL.' THETA AURORAFROM MULTISATELLITEOBSERVATIONS ET
asymmetryin the luminosityand electronfluxes acrossthe polar arc, whereasin the secondcasea low level of luminosity and precipitation is observedon each side of the polar arc. The first and third cases differ on the large-scale convectiondirectionin the vicinity of the arc; namely, in
the first case, convection reverses on both sides of the arc, whereas in the third case, convection is antisunward

17,431

a

3August 1986

8-

throughoutthe arc and on both sides [cf. Weiss et al.,
1993].

Examining Viking and DMSP auroralimagesas well as the corresponding particle data, Austin et al. [1993] and Makita et al. [1991] both reportedthe existenceof polar

I
17

I
18
UT

I
19

arcs thepoleward of thesoftprecipitation on edge region.
This region extendsfrom the dawn (dusk) sectorof the auroral oval precipitation for interplanetarymagnetic field
3 August 1986
nT

X-comp

h=-157'

(IMF) By< 0 (>0) in the northern (southern) hemisphere. Movement polararcsis alsocontrolled the IMF By of by sign:along Bydirection thenorthern the in hemisphere and
opposite to it in the southern hemisphere [Craven and Frank, 1991]. The asymmetryin 0 aurora locationrelative to the noon-midnight meridional plane in both hemispheres[Craven et al., 1991] or 0 aurora and associated plasmain the magnetotail lobe [Huang et al., 1989] suggest conjugacy with a mirror reversal between hemispheres. Such mirror conjugacycan be interpretedas the

100

SOP 66.7'
o

-100 DIX 68.2··,..·
lOO

IZV 70.3'

UED 71.8'
lOO

result of the IMF By influence on the location of

_·Oo

VIZ 73.6' reconnection regions between the interplanetarymagnetic lOO and geomagneticfield lines in the northern and southern -lOO hemispheres. 17 ' ' ' 1·8 ' ' ' 1·9 ' Section 2 containsa brief descriptionof satellitesand instruments and presentsdata about the precipitationmorphology.Discussion and interpretation follow in section3, Figure 1. (a) By andBz components the interplanetary of with the conclusions given in section4.

magnetic field (IMF) as measured IMP 8 on August3, by 1986. (b) Variations the geomagnetic x component of field alonga meridional chainof magnetic stations (intermediate 2. Instrumentation and Observations latitude [ILAT], geomagnetic longitude): Sopochnaya Viking plasmainstrumentation consists sevenspec- (SOP; 66.7Ü,157.4Ü); Dixon Island (DIX; 68.2Ü,155.8Ü); of
trometer units, of which three have been used herein.
electron flux over the en-

These measured the directional

Izvestiya Island (IZV; 70.3Ü,158.8Ü); Uedinenia Island (UED; 71.8Ü,158.8Ü); Uyze Island (VIZ; 73.6Ü,156.2Ü).

ergy range 0.01-200 keV and ion flux over the range 0.04-40 keV. A detaileddescription given by Sandahl is (x,y,zin Re). The By (B components from + 10 nT 0 vary et al. [1985]. The satellite moved along the orbit with (+8 nT) at the start of the aforementioned interval to 13,530-km apogeeand 817-km perigee at an inclination -10 nT (-2 nT) by the end of the interval. A transition of 98.8 Ü.
The DMSP satellites measure fluxes of electrons and

of theIMF By(Bz)components onesignto theoppofrom
site took place at 1750 (1800) UT. Figure lb presentsthe variation of the geomagnetic field x componentfor a chain of magneticstationsnear 2400 magneticlocal time (MLT) along 157 geomagnetic Ü meridian.A substorm expansion phasebeganat 2015 UT with a sharpdecrease the x component auroralzone of at stations (67-68Ü). indications a growthphasehad previof

ions within the 0.032-30 keV energyrange [Hardy et al., 1984]. Sincedetector apertures alwaysoriented are toward the zenith, only particleswell insidethe loss cone are observed the latitudes interest at of herein. One full spectrum is obtainedper second.Both the DMSP F6 and F7 satellites move in sun-synchronous, nearly circular orbits of
835-km altitude.

ously been observable about since 1800UT throughout the
latitudinalchain. Specifically,clear x component decreases were noted for stations located poleward of the auroral

2.1 Precipitation Structure as Observedby Viking
latituderegion on August3, 1986, between1710 and 1930 UT, which crossedhigh latitudesfrom the prenoonto postmidnight local time sectors.The location of auroral precipitation boundaries high latitudes closelyrelated at is

We considerfirst a Viking pass throughthe high- zone (70-72Ü), which coincides with the IMF southward
turning.During the Viking passan increase the AU and in AL indiceswas recorded. Thus the Viking passincludesa magneticallyquiet interval prior to 1800 UT and also the early stages a growthphaseup to 1930 UT (45 minutes of to interplanetary parameters, particularly Bz andBy before expansiononset). the components. Figure l a showsthe IMF variationsas meaUntil 1828 UT, Viking measured fluxes consistent with suredby the IMP 8 satellite,locatedat (32.5,-13.5,-15.4) theplasma mantle [Newell al., 1991] et corresponding to a


17,432

FELDSTEIN ET AL.' THETA AURORA FROM MULTISATELLITE OBSERVATIONS
beams. With the exceptionof an upward loss cone above 1 keV, the ion populationis isotropic,whereasthe electron populationhas a highly varying flux level and pitch angle characteristics.After 1906 UT, only faint and sporadic

furthest poleward position of 84.8 Ü invariant latitude (ILAT)/917 MLT. Thereafter, fluxes dropped below the

onecount readout per threshold x 10 ergscm s-· (6 -3 -2 sr-t.keV-· for electrons 5 x 10 erg cm s-· sr and -5 2 -·

electronbeams keV forions). region, was -1 This which devoidprecipiof
tation, is considered be the polar cap. Figure 2 showsa to Viking energy-time spectrogramfor electrons (top) and ions (middle) starting from 1830 UT, when the satellite was still inside the polar cap. Allowing -20 min lag time for the IMF observedby IMP 8 to the Viking observation
electron

appear,and from 1911 UT (73.6Ü ILAT! 0318 MLT) up to 1918 UT (69.4Ü ILAT/0302 MLT) the
flux was near instrument threshold. The ion fluxes

point (consisting propagation of timesof 7 min to the bow

subside as well, with their spectra becoming softer, and the ion beams disappear.For a detector with a comparatively high threshold, this region might be considered empty polar cap. However, detailed analysis of the ion

shows a flux of-3 x 10 ionscm s-· sr that 2 -2 -· shock, 4 min to the magnetopause, and 9 min delay in the spectra ions above keV stillex1 responseof geophysicalphenomena [Clauer and Banks, eV-· of isotropic withenergies 1986; Greenwald et al., 1990]), it is believed that for the ists. This flux is slightly above the spectrometer threshold data presented in Figure 2, the pertinent IMF conditions and indicates continuity of auroral plasma precipitation of are Bz < 0 and By < 0. Duringthe first part of the orbit, along the Viking pass,thoughcharacteristics fluxes may prior to -1841 UT, the ion detector experiencedso much vary substantially. noise in the channels between 1.2 and 40 keV that this After 1918 UT the intensity of auroral plasma fluxes portion of the data was not useful. begins to increase.In the interval until 1921 UT (67.5Ü At 1836 UT the satellite first recordedauroral plasma ILAT!0256 MLT) at the height of-8400 km, structured fluxes, namely electrons with energy 0.2-0.3 keV and bursts of soft electrons with E< 1 keV occur, but ion fluxes to 105e- cm s-· sr eV-· andionswithen- fluxes remain at low levels. Around 1921 UT a sharp inup -2 -· ergy0.5 keV andflux of about x 10 ionscm s-· sr 5 2 -2 -· crease of structured fluxes and energy of precipitating eV-·. Auroral plasma with such parameters observedelectronstakes place. At the polewardedge of this region, was until 1837 UT (85.0Ü ILAT/0728 MLT) when both the en- ion fluxes also increase, beginning with higher energies.
ergyandflux of electrons increased 0.5 keV and10 e- Such characteristics of this event as its location at the to 6 cm s-· sr eV-·, respectively. -2 -· Apparently, wasthe polewardboundaryof intense,structured it electronprecipitafirst encounterof Viking with the plasmaabove the 0 aurora. A 0 aurorawas distinctlydiscerned Viking images in up to 1828 UT when optical observations the northern in hemisphere ceased, whereasin the southern hemisphere the DE 1 satellite recordedaurorasup to -20 UT [Craven et
tion, and the latitudinalcharacterof the ion energydispersionjustify identificationas VDIS 2.

In the UT range 1921-1927 UT (63.2Ü ILAT!0242

MLT), Viking, at 7500-km altitude, intersected region the of harder electron precipitationwith electronenergy Ee up al., 1991]. to severalkeV, namely the discreteauroraloval. The highThe belief that this encounter was with the 0 aurora is energy electron channelsindicate that the stable trapping confirmed by calculating the motion of the 0 aurora fol- boundary,I,.·,which terminates outer radiationbelt and the lowing the imager observation.The directionof the motion which Feldstein and Galperin [1985] have argueddivides

of the luminosity bandis determined from the signof By the diffuse aurora from discrete, is encountered at 1927
[Craven and Frank, 1991], whereas the velocity of the 0 aurora in the southernhemisphere was observed allby sky camera [Feldsteinet al., 1992]. The resultingcomputation indicates that the Viking trajectory transversedthe 0 auroral band at -1837 UT. Figure 3 showsthe observed 0 aurora locations in the northern hemisphereat three UT. Fluxes of ions with Ei = 10 keV sharplyincreaseand are isotropicexceptfor the upwardion loss cone.At 1927 UT the ion distributionchangesfrom basically isotropicto
a double loss cone. After 1927 UT a characteristic de-

creaseof maximumenergyelectrons observed. is Magnetic field measurements show that the character of times' (1) 1756 UT, (2) 1819 UT, and (3) 1826 UT the field-alignedcurrentsalso changes the boundary at be[Elphinstone et al., 1993b]. The velocity of the 0 aurora tween HP and DP, as previouslynoted at other local times shiftduring thistime is -0.4 km s-· according all-sky [Poretara, 1977; Erlandson et al., 1991]. Poleward to camera observations Vostok station.The expectedloca- (equatorward) this boundary,the currentsare directed at of tion of the 0 aurora at 1837 UT is shownby "4," and the inward (outward) of the ionosphere, which corresponds to position of the Viking satellite at 1837 UT is given by the Region 1 (Region 2) currentsof Iijima and Poretara a cross. [1976]. In the equatorwardportion of the large-scaleinBoth the luminosity band and Viking move toward the ward current, small-scale but intense local outward fielddawn side from the noon-midnight meridian. This shift aligned currents were observed, corresponding indito and the complicatedstructureof the luminosityband incor- vidual discrete auroral forms. The boundary of bursty precipitatingelectronswith up porating three auroral arcs naturally explain multiple increasesin the auroral fluxes up to 1906 UT (75.9Ü ILAT! to tens of keV energy and decreasingintensity with in0325 MLT) seen in the spectrogram. Figure 3 also shows creasinglatitude occurredat -1921 UT. This is the back("5") the locationof the 0 auroraexpected basedupon re- ground boundary ('I'b) of energetic electron precipitation flecting the southern hemisphere observationsabout the and coincides with the auroral oval poleward boundary. noon-midnight meridian [Craven et al., 1991]. Between The location of the auroral oval polewardboundaryis simdeterminedbasedon both electrons 1837 UT and 1906 UT, Viking observedions with ener- ply and unambiguously giesfrom 1 to 10 keV and with a flux below103ions and ions. Though the intensityof ions in the polar cap is
cm-2 s-· sr-· eV -·. An increasein the electron flux is accompanied by low-energy, high-intensity upward ion
lower than in the auroral oval, it should be stressed that

significantion fluxes exist inside the polar cap in the vi-


FELDSTEIN ET AL.: THETA AURORA FROM MULTISATELLITE

OBSERVATIONS

17,433

VIKING DATA,
[KEY]
10

ORBIT

00896

DATE

860803
log [cts]

2.7

0.1

0.3

,i

log [cts]

3.0

1.7

0.1

0.3 { ,

i

{

UT

18:32

H(km)
I-LAT MLT

12875
85.1 08.53

18:36 12670 85.1 07.44

18:4o 12436 84.6 06.38

18:44 12174 83.8 05.43

18:48 11884 82.8 05.01

18:52 11564 81.5 04.29

18:56 11215 80.1 04.05

I
(KEY)
10

sP
Viking Data, orbit 00896 date 860803

mog (cts)
2.7

- 1.5

0.1

10

Log (cts)
3.0

- 1.7

0.1 I

0.3

180

0
I I I I

UT

H(km)
I-Lat MLT

19:02 10636 77.7 03:38

19:06 10212 75.9 03:25

19:10 9758 73.9 03:13

19:14 9274 71.8 03:04

19:18 8759 69.4 02:55

19:22 8215 66.8 02:48

19:26 7640 63.9 02:40

Figure 2. A spectrogram Viking data from pass 896 in the interval from 1830 UT to 1930 of UT, August3, 1986. Electrons at top, ions are in the middle,and pitch anglesare at the botare tom; the plottedcountrate is proportional the differentialenergyflux in arbitraryunits. to


17,434

FELDSTEIN

ET AL.: THETA AURORA FROM MULTISATELLITE

OBSERVATIONS

12

·
.

80'
5
I

70'

60'
06

polar arcs, there exists continuous precipitation.Most intense bursts take place over an interval 1830:45-1831:40 UT, with the maximumenergyflux at 1831:22 UT (72.7Ü
MLAT/0150 MLT) and at 1834:10-1834:42 UT, with a

peak energy flux value at 1834:13 (74.1Ü MLAT/2330 MLT). These bursty locations are marked by thick lines along the satellite trajectory. Within this interval some electronspectrashow signsof "monoenergetic peaks,"indicating accelerationby field-aligned potential drops. The energy flux and averageenergiesare lower during the second burst than in the first.

I

The region of structuredsoft precipitationis not continuousalong the satellite trajectoryfrom the dawn sector to the dusk sector. Over the interval 1834:57 UT (74.0 Ü MLAT/2305 MLT)-1836:15 UT (71.4 Ü MLAT/2210 MLT), very soft electron and ion precipitation without
bursts of electrons was observed. The character of fluxes

and their energy spectrain this region were similar to that describedabove from Viking data in the interval 19111918 UT. It seemslikely that the satellitesintersected the 24 soft precipitation regionfrom the plasmasheetperiphery. At 1838:05 UT the energy of the electronssharplyinFigure 3. Successive locationsof the central portion of the 0 aurorain Viking imagesfor (1) 1756 UT, (2) 1819 creased,and a region of hard precipitationwith high ion UT, and (3) 1826 UT. The positionat (4) 1837 UT is esti- fluxes in the range -10 keV is observed,which reachesto mated from all-sky camera observations. The positionat 1838:35 UT (66.5Ü MLAT/2100 MLT), the HP. The subse(5) 1906 UT is that of the mirror reflection about the quent decreaseof auroral electron flux and energy before noon-midnightmeridianof the 0 auroraobserved DE 1 1839:02UT (64.8ÜMLAT/2052 MLT) is a consequence by of in the southern hemisphere. The two crosses mark the Vi- the satellite'sentry into the diffuse precipitation region. king satellite location at 1837 and 1906 UT. In the morning and night sectors DMSP F7 crossed high latitudes. Over the interval 1836:27 UT (69.4 Ü MLAT/0648 MLT)-1839:41 UT (74.5 Ü MLAT/0449 cinity of the 0 aurora.The variabilityof plasmafluxesob- MLT), auroralenergy electronand ion precipitation inare served in the vast region poleward of the auroral oval tense and structured(MP in Figure 5). The location of the testifiesto the complicated structure the plasmaregion precipitationregion in the dawn sectorsuggests connecof a at the periphery of the plasma sheet in the magneto- tion with the boundaryregion of the magnetosphere the in spherictail. deepmagnetotail flanks [Feldsteinet al., 1994a]. Precipitation from 1839:41 UT to 1840:12 UT (74.6Ü 2.2 Precipitation Structure as Observed from DMSP F6 MLAT/0421 MLT) substantially differs from precipitation and F7 that occurreddirectly poleward; namely, auroral electrons are virtually absent,whereaslow-energy ions (0.1-1 keV) Figures4 and 5 presentthe spectrograms precipitat- are present. Their spectral characteristics akin to the of are ing electronsand ions by DMSP F6 and F7, respectively, Viking data from 1911 to 1918 UT and the F6 observaand Figure 6 showsthe satellitetrajectories polar pro- tions from 1834:57 to 1836:15 UT, which were classified in jection and auroral oval and the location of UV auroras as soft but unstructured precipitation. The locationof this basedon Viking and DE 1 imagesfrom 1814 to 1824 UT. high-latituderegion suggests probableorigin in the pea Figure 4 showsthat F6 crossedthroughthe high-latitude riphery of the magnetospheric plasma sheetregion, though night sector region from dawn to dusk. In the interval plasma spectralcharacteristics not unlike thoseorigiare 1826:30 (61.5 Ü MLAT/0425 MLT)-1827:49 UT (65 Ü natingin the plasmamantle. MLAT/0417 MLT), weak fluxes of soft electronswith inBeginningwith 1840:12 UT, fluxes of more energetic creasingenergy with increasinglatitude occur (DP). This electrons and ions appear,initially relativelyhomogeneous is the region of stabletrapping(the outer radiationbelt), in character but distinctly structuredfrom 1840:50 UT sinceany precipitation auroralenergyions is absent. of In (74.6Ü MLAT/0355 MLT) until 1847:34 UT (64.5Ü MLAT/ the interval 1827:49-1828:49 UT (67.5 Ü MLAT/0338 00:00 MLT). This latter intervalis the regionof soft strucMLT) the flux of precipitating electrons with severalkeV tured precipitation (SP). Three particularlystrongburstsof energy substantially increases, with suchflux constituting electronprecipitationwith accompanying intensification of luminosityalong the auroral oval (HP). Some of the elec- ion fluxes in the energyrange Ei- 1-10 keV take place between 1841:19 and 1842:34 UT. This interval is marked tron spectra this intervalshow signsof acceleration. in In the interval 1828:49-1838:05 (67.7 Ü MLAT/2112 by the thick line in Figure 6 on the DMSP F7 trajectory MLT), structured precipitation of soft electrons with and corresponds a 0 aurorafrom the Viking image.Parto Ee < 1 keV was observed,superimposed bursty elec- ticle precipitation on clearly demonstrates the "bar" in the that trons of higher energy associated with polar arcs as well 0 aurorahas a fine structure and actuallyconsists sevof as ion precipitationwith Ei · 1-10 keV (SP). Betweenthe eral polar arcs. The fine structureof the 0 aurora for Audawn sector auroral oval and the electron bursts above the gust 3, 1986, was first shown by Feldstein et al. [1992]


FELDSTEIN

ET AL.: THETA

AURORA

FROM MULTISATELLITE

OBSERVATIONS

17,435

F7
12

86/215

Electrons

Ions

Log JE
9 4-

Log E ave
Elec

Log
E Flux
Ion

II
_

I

10

,, i i Jit ·ti tl ·li.i l' ,. I
t

,i tl
c:

03

I,'. 2i ill,It·11t i·,tt .,'..
,jill
18:37:08 70.9 71.4 203.3 06:30

j
i

I
t I I II ,,
I ij ,ii
It iI

I !tl
t

- ,ti1! '
UT MLAT GLAT GLONG MLT

*t· Ii i1'
Iii

I ! illi j
I

Ill j I
·t i! 11tl 4

rill tJi "1' '
18:42:17 74.5 81.1 128.4 02:42

I,' ' ',' ,,* ,,',,l,!lij!i ,, t, I li l, I 'at'
I
i

't ' Iii 1 1

18:32:00 58.6 54.4 219.3 08:14

18:34:34 65.3 63.1 213.6 07:34

18:39:42 74.6 78.6 180.8 04:49

18:44:51 70.7 76.3 86.8 01:01

18:47:25 64.6 68.5 70.3 23:59

18:50:00 57.4 60.0 62.2 23:22

Aug 3

I

MP

I

SP

HP '·
DP

II

I

Figure 4. Spectrogram DMSP F6 particle observations the interval 1825 UT to 1843 UT. of in The spectrogram showsdifferentialenergy flux from 32 eV to 30 keV in units of eV/(cm·- sr s eV). The top line plot showstotal energyflux (eV/cm·- st), and the bottom line plot showsavs erage energy (eV). Note that the ion energy scaleis inverted.

basedon all-sky cameraobservations the southern in hemisphere.At 1847:34 UT the satellite intersected intense an auroral arc, judging from the sharp increaseof both electron and ion fluxes over a wide energy spectralrange of up to tens of keV (HP). The auroral arc equatorialboundary was intersected at 1848:04 UT (62.7 Ü MLAT/2350 MLT), and the arc locationis in good agreement with the Viking image. The hard precipitationregion comprisesa very narrow latitudinal interval, as is typical for intervals that follow an IMF southwardturning but are prior to a magnetospheric substormexpansiononset. Equatorwardof the hard precipitation, the spectrum auroralelectrons of becomes abruptly softer, with maximum energy decreasing with latitude and ion precipitationabsent.This is the outer radiationbelt region,which is characterized diffusepreby cipitation (DP). The burstsof electronprecipitationwith Ee > 1 keV are responsible for generatingpolar arcs at high latitudes in-

side the auroral oval. The SP region has apparentlya complicated structure with soft structureless precipitation at very high latitudes. It is possiblethat such a precipitation is connectedwith the plasma sheetperipheryin the magnetospheric tail. Passesby DMSP F6 and F7 show the existence of intense ion fluxes in the polar cap. Counter to the suggestions Troshichev of and Nishida [1992] no sharp dropoutof ions in the polar cap is observed. Thus the determination of the poleward auroral oval boundary based on ion observationsaffords no advantageover the more standard usageof electrondata.
3. Discussion

Observations the auroral plasma during an interval in of which the IMF turned southward but well before any substormonset show that precipitation in the dawn and night sectors occurs practically continuously inside the


17,436

FELDSTEIN

ET AL.: THETA

AURORA

FROM MULTISATELLITE

OBSERVATIONS

F6

86/215
Electrons

Log JE
9

_

Log E ave
2

Log
E Flux

Elec 10

Ion 8

.i.

1

ttl ti

..

I
ú

t!

t1
I
· f

I

I

I
1

I

I

II

,.

ú

Iil

·!
18:35:17 73.7 77.8 42.7 22:51

.1
18:37:51 68.4 70.4 21.7 21:16 18:40:25 60.9 62.0 12.3 20:22 18:43:00 52.3 53.3 6.7 19:51

5 L3

UT MLAT GLAT GLONG MLT

18:25:00 56.8 60.8 163.6 04:49

18:27:34 64.3 69.3 154.9 04:10

18:30:08 70.9 76.9 136.7 03:02
HP

18:32:42 74.6 81.1 91.9 01:07

Aug 3

I

DP

I

I

SP

I

HP DP

II

Figure 5. A DMSP F7 spectrogram the interval1832-1850UT on August 1986. for 3,

wide

interval

of latitudes

from

the auroral

oval

to 85 Ü.

alignedcurrents corresponds the widely known pattern to found by lijima and Potemra [1976], namely that in the dawn sectorthe inflow currentslie polewardof the outflow currents. The region of inflowing currentsis here widenedand shiftedfar to the poleward region.The largescale inward current region also containssmall-scalebut mean energy, currents,and plasmadensity.The energy intenseoutflowing currentsthat apparentlyare connected flux hasa sharp maximum 9 mW m-2 at theboundarywith precipitation of abovethe polar arcs. betweenDP and HP at 1927 UT and dropsoff from this If the particle detectors had had a relatively high value movingeither equatorward poleward. or Fluxeshave threshold(or were plotted on an insensitive scale),the dea local maxima of several milliwatts per square meter creaseof particlesin the interval from 1911 UT to 1918 above the polar arcs in the interval from 1837 to 1906 UT could be mistakenfor a discontinuity precipitation in UT. The regionwith polar arcsis distinctlydiscerned from between the auroral oval and the 0 aurora, although no the auroraloval regionby its intensityof bothelectronand such discontinuityexists, either in the electronor ion data. ion precipitation. interesting An fact aboutthe large-scale Figure 2 presents weak low-energyion precipitation the in field-aligned current distribution should be stressed, this region. In Figure 8, ion spectrafor the time interval namely that inside the auroral oval, current flows out of 1912:00-1912:40UT (one satellitespin period)are shown. the ionosphere, whereas the general in vicinityof the polar The spectrahave been binned and averagedfor the pitch

Thus for this particular case the full set of detailedobservations better correspondsto the hypothesisof plasma sheetwidening[Meng, 1981; Murphree and Cogger, 1981 ] than bifurcation[Frank et al., 1986]. However precipitation intensitymay vary by ordersof magnitude within the range of continuousprecipitation.Figure 7 presentsthe variations in the electron energy flux along the Viking passin the energy range 40 eV-30 keV, along with their

cap arcs, the large-scale currentflows into the ionosphere (of course,precisely above the arcs, small-scalecurrents
are upflowing). Such interrelated location of the field-


FELDSTEIN ET AL.: THETA AURORA FROM MULTISATELLITE
12

OBSERVATIONS

17,437

tudes mapping onto field lines on the plasma sheet,which are highly stretched [Sergeev et al., 1983; Lyons et al.,
1988].

Let us consider energy spectra for different structured plasma regions at Viking heights.The soft precipitationregion poleward of the 0 auroras deservesspecial interest, since it is presumablyconnectedwith the low-energy electron layer discovered Parks et al. [1992]. Figure 9 preby sentselectronand ion spectrain this region at 1836 UT.

For all pitch angles,electronfluxes sharplydecrease for
E > 1 keV, with maximum flux values observed at 200 eV

to 500 eV. The pitch angle distributionshownin Figure 9a is isotropicfor downgoingelectrons,but upgoingelectrons DMSP F6 have only approximately 1/4 the flux, meaning a net flux into the ionosphereexists. Their intensity is enoughfor luI minosity excitation in the several tens of Rayleigh of the atomic oxygen red line. The existenceof weak diffuse auroral luminosity poleward of a polar arc [Austin et al., 1993] is consistentwith the observed downward flowing 18.50 O0 soft electron fluxes. At 1836 UT ion spectraare available UT for E < 1 keV only becauseof interferencefrom other instruments.Fluxes of ions are approximately 2 orders of Figure 6. Trajectoriesfor the DMSP F6 and F7 and Vi- magnitude less than electron fluxes. For_both ions and king spacecrafton August 3, 1986, in correctedgeomag- electrons, strong anisotropy is observed. The ion fluxes netic coordinates. The start and end UT times are listed along the magnetic field lines flow into the ionosphere, along with the major plasma structures:DP, diffuse preThe inward cipitation; HP, hard precipitation; SP, soft precipitation; with little return flow out of the ionosphere. ion flow conforms with Parks et al.'s [1992] measureMP, morning precipitation; PC, polar cap precipitation. The burstsof auroral precipitationalong the DMSP F6 and ments of earthwardflowing ions in the LEL at a distance F7 trajectories are marked by heavy lines. The dashed of--15 Re. However, in contrastto the Viking electronoblines mark the medium position of auroral luminosity re- servations, Parks et al.'s [1992] measurementsshow elecgion based on Viking images between 1814 and 1824 UT tron fluxes in the LEL directed toward the tail. This find(data courtesy of R. Elphinstone).The solid line is the ing means that either the sourceof auroral electronsin the projectionof the southernhemisphere polar arc basedon a LEL is located at heights between 2 Re and 15 Re, or if DE 1 image at 1825 UT into the northernhemisphere. the sourcelies in the magnetotail,a searchfor suchfluxes at the periphery of the plasma sheet is appropriate.The angle intervals 10-15 Ü 15-30 Ü 75-105 Ü and 150-170 Ü. availability of precipitatingsoft electronsat -12,500-km , , , Spectrafor 10-15 Ü and 150-170Ü intervals correspond to height removes difficulty when attempting to connect the smallestand largest pitch angles available for this or- Parks et al.'s [1992] LEL with luminositypoleward of the bit. Ion fluxes are isotropic over pitch angles and slightly polar arcs region. above detector threshold. Such isotropic distribution is Let us considerthe electron and ion spectraabove the characteristic for ion fluxes above the ionosphereat lati- polar arcs. Figure 10 presentsa typical spectra observed

?·4 18.25 OP UT

10

Viking 3 August1986 Orbit 896 V3 Moments
x

5.0 ..·.·._,,w,..·· ................................................................... 2.0 ........ · ·.q
1.o o ......
-2.0

....... .... :-:' :.......

..... ;............................ ....

18.30

18.42

18.54
UT

19.06

19.18

19.30

Figure 7. Summary plot of Viking data from orbit 896, August 3, 1986. The plot shows the electron energy flux, field-aligned currents(downward positive and upward negative), average electron energy, and electron density for the electronenergy range 40 eV to 30 keV.


17,438
Orbit
10 6

FELDSTEIN ET AL.: THETA AURORA FROM MULTISATELLITE
896 ions 19'12'2019:12:40

OBSERVATIONS

sharpcutoff and virtual absence ions with energy above of

1 keV. The spectra indic that at the satellite ate heightof

11,000 a field-aligned km electric fieldexists accelerthat
ateselectrons toward the Earth and ions in the oppositedirection. It also leads to the appearance a maximum in of the electronspectrum the energyof severalhundredeV. at For the auroral oval latitudes, electron spectra,presentedin Figure 11 (1922' 14 UT, 8200 km), differ only slightly from the spectra above the polar arcs; namely, electronsare field-aligned,with a peak in the 200-400 eV range. At the latitudesof diffuse precipitation region near the equatorial oval boundary, as shown in Figure 12 (1928:08-1928'18 UT), the electron spectrumis isotropic

105

ú ',:..
·
m.., .......15-30

104

X

:=)

10

3

10-15

Pitch;'·i:i·· angle Time ,UT, · .,.:·
19'12:30-32
19-12:28-30

\ ·:;·
ú ·.

in the upperhemisphere, maximumat several the hundred eV disappears, and the spectrumbecomessmoother.Ions with pitch anglesaround90Ü prevail.
Observations of DMSP F6 and DMSP F7 satellites al-

..... 75-1 ·'" 05
102
.01

....'*"-' 150-170
..................
.1

19:12:34-35
19:12:20
I
10

'%.,;,.
-,·
100

ENERGY (keY)

Figure 8. Ion energy spectrafor the region betweenthe auroraloval and polar arcsfor differe PitchanglesdurTM ing a 20-s interval of the satelliterevolutionfrom 1912:20
to 1912:40 UT.

from Viking (1900:23 UT). The fluxes of precipitating electrons peak at smallpitch angles, with intensity reach-

low a comparisonof the spectraof electronsabove polar arcs at different heights.Figure 13 showsan electronspectrum during a burst, as indicatedby the thick lines on Figure 6. The sharp maximum at several keV is a monoenergetic peak, generally interpretedto imply a potential differenceof severalkeV above the ionosphere. A comparison with the Viking spectrogramsdemonstrates that most of the potentialdrop must lie below 2 Re. Figures 13a and 13b presenttwo spectraof DMSP F6 at 1831:22 and 1834:13 UT. The first one corresponds to the positionof the polar arc, as recordedin the image of

-2 -· ing 106e- cm s-] sr eV-] at 200 eV. Thefluxes -2 -· re- Viking, with an energyflux of 10·Ü eV cm s-1 sr mainsubstantialhigher at energies well,about at eV-] with a peakof 3.2 keV. Thispolarcaparcwasalso as 105 1 keV, andgradually decrease 103at 6 keV. Above1 observedby DMSP F7, as shown in Figures 13c and 13d to
The F7 spectra keV upgoingelectrons are virtually absent. The spectra (at 1841:32 and 1842:34 UT, respectively). ions inside and outsidethe loss cone are practicallyidenti-

cal, with the fluxessubsiding from-103 ionscm s-· -2

show signs of field-aligned acceleration 1 keV and 1.4 to keV, respectively.The secondarc (Figure 13b) also resr-· eV-· at 40 eV to -2 x 10· at 10 keVi A typical veals signsof field-alignedelectronacceleration only but

upgoing beamis characterized a sharp ion by increase of

flux up to 104ionscm s-] sr eV-· at 40 eV, witha -2 -·
Viking
10 9
ú ' .......

to 250 eV, withan intensity only3.1 x 10 eV cm s-· of 9 2 sr eV-·. Thissecond -1 enhancement energy inof the flux
Viking ions 18:36:30107
Time
-18:36:30

electrons
i .......

18:36:30i ........

18:36:40
i ........

18:36-40
(UT) Pitch
5-6

Time (UT)Pitch
; ......e ......
.... -·-.--

angle

angle

18:36:31 18:36:35
18:36:40

i 0-21 99- 105
I 68-1 74

.......· ......

18:36-35

88-94

·0 8
107

106

· .... 18:36:40 7.-. 172-174
level ·

·..
105

· 10 ·

ú .... ·

\/

10 s
.01 .1 I 10 100

104
.01

.1

I

10

ENERGY (keV)

ENERGY (keV)

Figure 9. Electron and ion spectraobservedby Viking in the region of weak diffuse precipitation poleward of a polar arc at 1836 UT. The actual time of each spectrum the one given unis der UT inside the frame. The presented spectrawere measured closeto the 0Ü, 90Ü, and 180Ü as as possible.


FELDSTEIN ET AL.' THETA AURORA FROM MULTISATELLITE OBSERVATIONS
Viking
109

17,439

electrons

19:00:23

- 19:00:33
107 ,
....

Viking
·. .... r

ions 19:00:23
Time
-..... a- .... .... +.-.

- 19:00:33
(UT) Pitch
8-9 88-93 169-171

angle

19:00:23 19:00:28 19:00:32

:' 10 8
· 107

,
106

One count level

\

·

10 6
105

·

105
104
.1 I I0 I 00

104
.01

.....
100

.01

ENERGY (keV)

ENERGY (keV)

Figure 10. Electronand ion spectra observed Viking abovepolar arcsat 1900 UT. by

put into the ionosphere was not revealedin the Viking images until 1828 UT. However, the location of the second burst in invariant latitude-local geomagnetictime coordinatesin Figure 6 preciselycorresponds the 0 aurorapoto sition recorded at 1824 UT by the DE 1 satellite in the southern hemisphere [Craven et al., 1991]. The correspondenceis "precise"in the sensethat no shift relative to the dawn-dusk plane is observed. The absenceof noticeableluminosityin the Viking image in the position of the secondelectronburst is due to both the lower fluxes and the softer spectrumof precipitating electrons above this second arc. According to Elphinstoneet al. [1993b] the Viking UV camera is less

responsive fluxes of precipitatingelectronswhose enerto gies are below about 500 eV. In any event, the 250-eV average energy electronsabove the secondpolar arc are absent from the Viking images. This finding suggests that, for reliable determinationof the polar cap boundary, Viking imagesshouldbe supplemented particle data. by
4. Conclusions

Viking electrons

19:22-14-

19:22:25

lO 9

·
;n

1Ü 8
lO 7

l _ :::::::
Pit·_h

The data presentedabove suggestthe following interpretation.Polar arcs observedby Viking and DE 1 images presented Craven et al. [1991] reflect only the positions by of the most intense and energetic electron precipitation. Using particle detectorswith sensitivityto lower flux inflows and apparentlylower characteristic energies,a more complex luminosity structure is revealed. The conclusion of Craven et al. [1991] about the asymmetrical locationof the 0 aurora in both hemispheres August 3, 1986, (in on the morning sectorin the northernhemisphere and evening sectorin the southern hemisphere) applicableonly to the is most intenseluminositysites (which also represent harder electron precipitation). Precipitation observationssupport the location of the most intensepolar aurorain the northern hemisphereas being in the morning sectorbut show that a less intense arc occurs in the evening sector, in a conjugatepositionto the 0 aurora observedin the southern hemisphere.Thus the precipitationpattern, but not intensity, is conjugate.The asymmetryin polar arc intensities is

apparentlycontrolledby the IMF By component. is It known that duringintervalswith By < 0 the polar cap

boundary in the dawn (dusk) sector shifts poleward (equatorward)[Elphinstone al., 1990]. It is possiblethat et 104 the apparentshift is at least partially the consequence a of .01 .1 I I0 I 00 relative luminosity decreasebut not actual disappearance. ENERGY (keV) The conjugacyof the polar arcs location (althoughnot intensity)in geomagnetic coordinates and the continuityof Figure 11. Electronspectra observed Viking at auroral precipitation between the auroral oval and the polar arcs by oval latitudes at 1922 UT. both suggest that field lines above the 0 auroraare closed


17,440

FELDSTEIN ET AL.: THETA AURORA FROM MULTISATELLITE

OBSERVATIONS

Viking
109

electrons

19:28:08,

19:28:18
108

Viking ions 19-28:08

- 19:28:18
Pitch
85-91

................

.................

Time(UT)
19:28:13

angle

10e

107

..,·!·,,,a, ' · 19:28:18 9-10
19:28:08 169

107

10 6

One ntle c
10 6
X

O/·U vel·
19:28:18 9

·'*' \
105
X

-,

',

'·l..

One nt!e c
I 04

o·u ve·""\ \ x · \a'""·
.1 I I0 I 00

ú J

105

--

Time Pitch (UT) angle·

·.

..i u.

104
.01

.... 19:28:08 +'-' .......................
.1 I 10 I 00

10 3
.01

ENERGY (keY)

ENERGY (keV)

Figure 12. Electron and ion spectraobservedby Viking in the diffuse precipitationregion equatorward the auroraloval at 1928 UT. of
and support the hypothesisof plasma sheet widening or LLBL expansionrather than plasma sheet bifurcation to explain 0 aurora.However, because electricfield observations for this event [Feldstein et al., 1994b] and previously reported0 aurora [Frank et al., 1986] show that convection reversals occur on both sides of the 0 aurora, we beSome other conclusions can also be reached from our

study. Detailed examinationof the precipitationcharacteristics immediately poleward of the main auroral oval indicates a correspondence between this region (SP or PDA) and the LEL found at high altitudes by Parks et al. [1992]. Finally, our observations indicate that, contraryto

lieve that only the hypothesisof an expandingplasma previoussuggestions [Troshichev and Nishida, 1992], ions sheet adequately explainsall observations. The causeof are present polewardof the auroraloval duringa 0 aurora

theBy-dependent asymmetry theintensity theauroral event, do notserve better demarcate auroral in of and any to the
precipitation still unresolved. is oval boundaries than do electrons.

Electronenergyflux
10

Electronenergyflux
i . i . i . i ·'1'. i , i , i". i . i'·'1". i . i . i '. i

(a)

,

(b)

.

·

7

1.1.1,1.1.1,1.1.1.1.1.1,1,1.1

(c)

ú .·1'.. i .'1 i©.. i . i . i . i . i . i . ii .i

Log average energy (eV)

Figure 13. Electronspectra abovethe polar arcson August3, 1986, observed the satellites by
DMSP F6 at (a) 1831:22 UT and (b) 1834'13 UT and DMSP F7 at (c) 1841:32 UT and (d)
1842:34 UT.


FELDSTEIN ET AL.: THETA AURORA FROM MULTISATELLITE OBSERVATIONS

17,441

Acknowledgments. The authorsthank the refereesfor their Erlandson, R. E., D. G. Sibeck, R. E. Lopez, L. J. Zanetti, and T. A. Potemra, Observationsof solar wind pressure unusuallythoroughand helpful suggestions. Discussions with initiated fast mode waves at geostationary orbit and in the B. Hultqvist, J. Craven, and Y. Galperin have been useful. polar cap, J. Atmos. Terr. Phys., 53, 231-239, 1991. R. Elphinstone kindly supplied the Viking images, and O. Troshichev supplied the magnetic meridian observations. Feldstein, Y. I., and Y. I. Galperin, The auroral luminosity structurein the high-latitudeupper atmosphere: dynamIts This researchhas been supportedby the Russian Foundation ics and relationship to the large-scale structure of the for FundamentalResearchat IZMIRAN by grant 93-05-8722 Earth's magnetosphere, Rev. Geophys., 23, 217-275, 1985. and at PGI by grant 94-05-16273a and by the International Science Foundation grant M6P000. Viking was financed by Feldstein, Y.I., and Y. I. Galperin, An alternative interpretation of auroral precipitation and luminosity observations the Swedish Board for Space Activities and managedby the from the DE, DMSP, AUREOL, and Viking satellites in SwedishSpaceCorporation. Work at JHU/APL was supported terms of their mapping to the nightsidemagnetosphere, J. by NSF GEM grant ATM-9300866. The Editor thanks two referees for their assistance in Atmos. Terr. Phys., 55, 105-121, 1993. Feldstein, Y. I., V. G. Vorobjev, S. V. Leontyev, R. D. evaluatingthis paper. Elphinstone,I. I. Alexeev, and E. S. Belenkaya, Auroras in the polar cap, IRF Sci. Rep. 209, Swedish Institute of Physics,Kiruna, 1992. References Feldstein, Y. I., R. D. Elphinstone, D. J. Hearn, J. S. Murphree, and L. L. Cogger,Mapping of the statisticalauAkasofu, S.-I., C.-I. Meng, and K. Makita, Changesof the roral distribution into the magnetosphere,Can. J. Phys., size of the polar cap during substorms(abstract), EOS 72, 266-269, 1994a. Trans. A GU, 72(44), Fall Meeting Suppl., 400, 1991. Austin, J. B., J. S. Murphree, and J. Woch, Polar arcs: New Feldstein, Y. I., A. E. Levitin, L. I. Gromova, L. A. Dremuhina, L. G. Blomberg, P.-A. Lindqvist, and G. T. results from Viking UV images, J. Geophys. Res., 98, Marklund, Electromagnetic weather at 100 km altitude on 13,545-13,555, 1993. 3 August 1986, Geophys. Res. Lett., 21, 2095-2098, Clauer, C. R., and P.M. Banks, Relationshipof the interplan1994b. etary electric field to the high-latitudeionospheric electric Frank, L. A., and J. D. Craven, Imaging resultsfrom Dynamfields and currents, J. Geophys. Res., 91, 6959-6971, 1986. ics Explorer 1, Rev. Geophys., 26, 249-283, 1988. Craven, J. D., and L. A. Frank, Diagnosisof auroral dynam- Frank, L. A., et al., The theta aurora, J. Geophys.Res., 91, 3177-3224, 1986. ics using global auroral imaging with emphasison largescaleevolutions,in Auroral Physics,edited by C.-I. Meng, Galperin, Y. I., and Y. I. Feldstein, Auroral luminosity and its relationshipto magnetospheric plasmadomains,in AuM. Rycroft, and L. A. Frank, pp. 273-288, Cambridge roral Physics, edited by C.-I. Meng, M. Rycroft, and University Press,New York, 1991. L. A. Frank, pp. 207-222,Cambridge University Press, Craven,J. D., J. S. Murphree,L. A. Frank, and L. L. Cogger, New York, 1991. Simultaneous optical observations transpolar of arcs in the two polar caps with DE 1 and Viking, Geophys. Res. Lett., Greenwald, R. A., K. B. Baker, J. M. Ruohoniemi, J. R. Dudeney, M. Pinnock, N. Mattin, J. M. Leonard, and 18, 2297-2300, 1991. R. P. Lepping, Simultaneous conjugate observations dyof Eastman, T. E., L. A. Frank, W. K. Peterson, and namic variationsin high-latitudedaysideconvectiondue to W. Lennartson, The plasma sheet boundary layer, changesin IMF By, J. Geophys.Res., 95, 8057-8072, J. Geophys.Res., 89, 1553-1572, 1984. 1990. Eather, R. H., Latitudinal distributions auroral and airglow of Gusev, M. G., and O. A. Troshichev, Simultaneous groundemissions:The "soft" auroral zone, J. Geophys.Res., 74, based observations polar cap arcs and spacecraftmeaof 153-158, 1969. surements particle precipitation,J. Atmos. Terr. Phys., of Eather, R. H., and S.-I. Akasofu, Characteristics polar cap of 54, 1573-1591, 1992. auroras,J. Geophys. Res., 74, 4794-4798, 1969. Eather, R. H., and S. B. Mende, High latitude particle pre- Hardy, D. A., L. K. Schmitt, M. S. Gussenhoven, F. J. Marshall, H. C. Yeh, T. L. Shumaker, A. Hube, and J. cipitation and source regions in the magnetosphere, in Pantazis, Precipitatingelectron and ion detectors(SSJ/4) Magnetosphere-Ionosphere Interactions, edited by for the block 5D/flights 6-10 DMSP satellites:Calibration K. Folkestad,pp. 139-154, Universitatforlaget, Oslo, 1972. and data presentation, Rep. AFGL-TR-84-0317, Air Force Eliasson, L., R. Lundin, and J. S. Murphree, Polar cap arcs Geophys.Lab., Bedford, Mass., 1984. observedby the Viking satellite, Geophys.Res. Lett., 14,
451-454, 1987.

Hones, E. W., J. D. Craven, L. A. Frank, D. S. Evans, and

Elphinstone, D., and D. J. Hearn, The auroraldistribution R. and its relation to magnetospheric processes, Adv. Space
Res., 13, 17-27, 1993.

P. T. Newell, The horse-collaraurora: A frequent pattern of the aurora in quiet times, Geophys. Res. Lett., 16, 3740, 1989.

Huang, C. Y., J. D. Craven, and L. A. Frank, Simultaneous observations a theta aurora and associated of magnetotail plasmas, Geophys. J. Res., 94, 10,137-10,143, 1989. By dependencies observed the Viking satellite,J. Iijima, T., and T. A. Potemra, The amplitudedistributionof as by field-aligned currentsat northern high latitudes observed Geophys.Res., 95, 5791-5808, 1990. by TRIAD, J. Geophys. Res., 81, 2165-2174, 1976. Elphinstone, D., J. S. Murphree,D. J. Hearn, W. Heikkila, R. M. G. Henderson,L. L. Cogger, and I. Sandahl,The au- Lyons, L. R., J. F. Fennell, and A. Vampola, A generalassociation between discreteaurorasand ion precipitationfrom roral distribution and its mapping according to substorm the tail, J. Geophys.Res., 93, 12,932-12,940, 1988. phase,J. Atmos. Terr. Phys., 55, 1741-1762, 1993a. Elphinstone, R. D., D. J. Hearn, J. S. Murphree, L. L. Lundin, R., L. Eliasson, and J. S. Murphree, The quiet-time aurora and the magnetospheric configuration,in Auroral Cogger,M. L. Johnson, and H. B. Vo, Some UV dayside Physics, edited by C.-I. Meng, M. J. Rycroft, and L. A. auroral morphologies, in Auroral Plasma Dynamics, Frank, pp. 177-194, Cambridge University Press, New Geophys.Monogr. Ser., vol. 80, edited by R. L. Lysak, York, 1991. pp. 31-45, AGU, Washington, D.C., 1993b. Elphinstone,R. D., K. Jankowska, S. Murphree, and L. L. J. Cogger, The configurationof the auroral distributionfor interplanetary magneticfield Bz northward,1, IMF Bx and


17,442

FELDSTEIN ET AL.: THETA AURORA FROM MULTISATELLITE

OBSERVATIONS

getic protons in the magnetotailcurrent sheetas the dominant source of their isotropic precipitationinto the nightnent, J. Geophys.Res., 96, 14,085-14,095, 1991. side ionosphere, Planet. SpaceSci., 31, 1147-1155, 1983. Meng, C.-I., Polar cap arcs and the plasma sheet, Geophys. Stasiewicz,K., Topology and positionas a function of interRes., Lett., 8, 273-276, 1981. planetary magnetic field and magnetic activity: CompariMizera, P. F., D. J. Gorney, and D. S. Evans, On the son of a model with Viking and other observations, J. Geophys.Res., 96, 15,789-15,800, 1991. conjugacyof the aurora:High and low latitudes,Geophys.

Makita, K., C.-I. Meng, and S.-I. Akasofu, Transpolarauro-

ras, their particleprecipitations, the IMF By compoand

Res. Lett., 14, 190-193, 1987.

Troshichev, O. A., and A. Nishida, Pattern of electron and

Murphree, J. S., and L. L. Cogger, Observedconnections between apparent polar cap features and the instantaneous diffuse auroral distribution, Planet. Space Sci., 11, 11431149, 1981.

Murphree, J. S., S. Ismail, L. L. Cogger, D. D. Wallis, G. G. Shepherd,R. Link, and D. M. Klumpar, Characteristics of optical emissions and particle precipitation the polar cap in arcs, Planet. Space Sci., 31, 161-172, 1983. Newell, P. T., W. J. Burke, C.-I. Meng, E. R. Sanchez,and M. E. Greenspan, Identification and observationsof the plasma mantle at low altitudes,J. Geophys. Res., 96, 3545, 1991.

ion precipitationin northernand southern polar regionsfor northward interplanetary magnetic field orientations, J. Geophys.Res., 97, 8337-8354, 1992. Weiss, L. A., P. H. Reiff, R. V. Hilmer, J. D. Winningham, and G. Lu, Mapping the auroral oval into the magnetotail using Dynamics Explorer plasma data, J. Geomagn.
Geoelectr., 44, 1121-1144, 1992. Weiss, L. A., E. J. Weber, P. H. Reiff, J. R. Sharber, J. D.

Obara, T., and L. aligned Viking, Parks, G. plasma
1992.

M. Kitayama, T. Mukai, N. Kaya, J. S. Murphree, L. Cogger, Simultaneous observations of sunpolar cap arcs in both hemispheres Exos C and by Geophys.Res. Lett., 15, 713-716, 1988. K., et al., Low-energy particle layer outside the sheet boundary,J. Geophys.Res., 97, 2943-2954,

Winninham, F. Primdahl, I. S. Mikkelsen, C. Seifring, and E. M. Wescott, Convection and electrodynamic signatures in the vicinity of a sun-aligned arc: Resultsfrom the polar accelerationregions and convectionstudy (Polar ARCS), in Auroral Plasma Dynamics, Geophys.Monogr. Ser., vol. 80, edited by R. L. Lysak, AGU, pp. 69-80, AGU, Washington, D.C., 1993. Zelenyi, L. M., R. A. Kovrazkin, and J. M. Bosqued,Velocity dispersed ion beams in the nightside auroral zone: AUREOL 3 observations,J. Geophys.Res., 95, 12,11912,139, 1990.

Peterson, W. K., and E. E. Shelley, Origin of the plasmain a cross polar cap auroral feature (theta aurora), J. Geophys.
Res., 89, 6729-6736, 1984.

Potemra, T. A., Large-scale characteristicsof field-aligned currents determined from the TRIAD magnetometerexperiment, in Dynamical and Chemical Coupling Between the Neutral and Ionized Atmosphere,edited by B. Grandel and J. A. Holtet, pp. 337-352, D. Reidel, Norwell, Mass.,
1977.

Y. I. Feldstein, IZMIRAN, Akademgorodok, Troitsk, Moscow Region, 142092 Russia. S. V. Leontjev and V. G. Vorobjev, Polar Geophysical Institute,Apatity, Russia. P. T. Newell, Applied Physics Laboratory, JohnsHopkins University, JohnsHopkins Road, Laurel, MD 20723-6099. (email: newell@oval.jhuapl. edu) I. Sandahland J. Woch, SwedishInstituteof SpacePhysics,

Sandahl,I., R. Lundin, and L. Eliasson,The hot plasma spec- P.O. Box 812, Kiruna S-98128, Sweden. trometers Viking, KGI Tech.Rep., 035, Kiruna Geophysical on
Institute, Kiruna, Sweden, 1985.

Sergeev, V. A., E. M. Sazhina, N. A. Tsyganenko, J. A. Lundblad, and F. Soraas, Pitch-angle scattering of ener-

(ReceivedJuly 8, 1994; revisedJanuary27, 1995; accepted January27, 1995.)