Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.izmiran.rssi.ru/magnetism/Feldstein_2012/PDF/375.pdf Äàòà èçìåíåíèÿ: Fri Mar 6 10:18:22 2009 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 21:03:39 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: iapetus

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 101, NO. A5, PAGES 10,737-10,748,MAY 1, 1996

Morphology of nightside precipitation
Patrick Newell,· YashaI. Feldstein,Yuri I. Galperin,andChing-I.Meng· T. 2 3
Abstract. Considerable informationon the stateof the magnetosphere embedded the is in structure nightside of charged particleprecipitation. reduceambiguityand maximize To the geophysically significantinformationextracted, detailedscheme quantitatively a for classifying nightside precipitation introduced. is The proposed system, which includes operationaldefinitionsand which has been automated, consists boundary1, the "zeroof energy"convection boundary (often the plasmapause); boundary the point wherethe 2e, large-scale gradientdEe/dXswitches from positiveto _<0(the startof the main plasma sheet);boundary2i, the ion high-energy precipitation cutoff (the ion isotropyboundaryor the startof the tail currentsheet);boundaries 3a,b, the most equatorward and poleward

electron acceleration events (spectra with "monoenergetic peaks")above0.25 erg/cms; 2
boundary4s, the transitionof electronprecipitation from unstructured a >10-km spatial on scale(spectra have 0.6-0.95 correlation coefficients with neighbors) structured to (correlation coefficientusually0.4 and below); boundary5, the polewardedge of the main auroraloval, markedby a spatiallysharpdrop in energyfluxes by a factor of at least4 to levelsbelow thosetypical of the auroraloval; and boundary the polewardedgeof 6, the subvisualdrizzle often observed polewardof the auroraloval.
originatingfrom the more stretched field line region where most electron acceleration events occur. This latter they Thirty years ago the first precipitation classification termedthe "boundary plasmasheet"(BPS). WYAH75 did scheme was introduced: Johnson et al. [1966] identified a not give clear operational definitionsfor identifyingtheir hard zone and a soft zone, the latter definedby the pres- CPS and BPS precipitationregions,choosingto illustrate ence of counts in the >80-eV electron detector but not in with examples.This has led to a certain ambiguityin terthe >21-keV detector.This classification systemhad an el- minology which further complicatesa discussion physof egant simplicity and operationalclarity not subsequently ics questions.For example, various interpretations the of approached. lacked only utility and geophysicalsignifi- poleward boundary of "Winningham's CPS" exist in the It cance. Burch [1968] showed that the soft zone existed on literature,including the start of structured satelliteauroral both the dayside and nightside but was softer on the imagery [Lui et al., 1977], the onset of highly structured former. Later it was shown that some of the daysidepre- electronprecipitation [Newell et al., 1991c], and the stable cipitation originatesin the magnetosheath [Heikkila and trapping boundary of >35-keV electrons[Deehr et al., Winningham,1971; Frank, 1971]. In a seriesof paperswe 1976; Weiss et al., 1992]. have previously discussed the classificationand quantitaWYAH75 associated the CPS with diffuse aurora, as tive identificationof varioustypes of daysideprecipitation definedby Lui and Anger [1973], and the BPS with dis[Newell and Meng, 1988a; Newell et al., 1991a, b; cf. crete aurora.They reportedthat the CPS was relatively unCambou and Galperin, 1974; Sauvaud et al., 1980; changedby the substorm cycle but that the BPS in quiet Galperin et al., 1976]. times had a deactivated phasein which only soft electron Twenty years ago Winninghamet al. [1975] (hereafter precipitation was observed as well as an activated referred to as WYAH75) made a major advance in de- (substorm) phase in which the BPS was energized.Later scribingthe nightsideprecipitation morphologyand its de- the view that the BPS was associatedwith the highpendenceon the substorm cycle. They divided the precipi- altitudeplasmasheetboundarylayer (PSBL) signature betation into plasmamappingto the near-Earthquasi-dipolar came common. For example, Eastman et al. [1984, field lines, which they referred to as "the central part of p. 1554] state that the BPS is the "probablelow-altitude the plasmasheet"and hencetermedthe CPS, and plasma signature of the [high-altitude] plasma sheet boundary layer" (see also Rostokerand Eastman [1987]).

1. Introduction and Background

·The Johns HopkinsUniversity AppliedPhysics Laboratory, as FG85) arguedagainstthis interpretation severalreafor Laurel, Maryland. sons. Some of their objections, when well understood, 2Ionosphere Radiowave and Propagation, Institute Terrestrial of prove to be primarily issuesof terminology.For example, Magnetism,Troitsk, Russia. 3Institute SpaceResearch the Academy Sciences of of of of FG85 did not believe that the region of hot plasma on quasi-dipolarfield lines earthwardof the tail current sheet Russia, Moscow. shouldbe termed the "centralplasma sheet";indeed, they argued that this term should apply only to the plasma that Copyright1996 by theAmericanGeophysical Union. lies tailward of the earthward edge of the current sheet. The apparent conflict between FG85 and WYAH75 was Papernumber 95JA03516. 0148-0227/96/95 JA-03516509.00 worsenedbecauseof the operationalambiguityin defining
10,737

Feldsteinand Galperin [1985] (hereinafter referredto


10,738

NEWELL ET AL.: MORPHOLOGY OF NIGHTSIDE PRECIPITATION

the CPS. However, even beyond all these (and other) issuesof terminology,FG85 had an unassailable argument in that the region identifiedas PSBL at high-altitude was only a tiny fraction of the full plasmasheetthickness; yet the low-altitudeBPS could well constitute majority of the the auroral oval width. The argumentsof FG85, further advanced Galperin and Feldstein[1991] with a discusin sion of recently discovered phenomena such as velocitydispersed ion signatures (VDIS-2), made clear to all who studiedthe issuethat in fact the PSBL does not map to the entire low-altitudeBPS. Instead,as originallyproposed by Feldsteinand Starkov [1970] and confirmedby many

others [Deehr et al., 1976; Weisset al., 1992], discreteauroral arcs map to the main plasma sheet which lies poleward(tailward) of the stabletrappingboundaryof 35keV electrons. FG85 also pointedout that an auroralclassificationschememust include an additionalprecipitation region, namely, subvisualdrizzle which lies polewardof the main auroral oval. They cited a long but little noted historyof measurements requiringsucha separate classification (e.g., Eather [1969]). In the presentpaperwe revisit the question nightside of morphologyfor three reasons: Much new research (1) has been done in recentyears on the phenomenology nightof side precipitation. This research makesit possible introto duce a nightsideclassification systemwith more geophysi- this association [Galperin et al., 1977; Jorjio et al., 1978; cally significant information. (2) Because it has recently Horwitz et al., 1986; Sauvaud et al., 1983]. However, the become obvious that researchers use the same terms in zero-energy boundaryobserved any given time doesnot at quite different ways, we wish to introduce operationally necessarily representa steady-state convection boundary, unambiguousalgorithms for identifying these significant sincemagnetotail convection highly dynamic[Mauk and is boundaries.(3) The phenomenology nightsideprecipita- Meng, 1983; Galperin et al., 1975]. Indeed, the lowof tion is more complexthan reportedby WYAH75, and, in energyequatorward precipitation ion often contains plasma someinstances, their discussion requiresmodification. No- originallyof ionospheric origin,apparently injected high at tably neither the CPS, however defined,nor any other pre- latitudes often in associationwith auroral arcs [Jorjio et cipitation region is substantially independent substorm al., 1985; Bosquedet al., 1986; Cambouand Galperin, of cycle. Indeed, the CPS of WYAH75 almost disappears af- 1974, 1982]. However, its subsequent convectiontoward lower latitudes is the result of ExB drift effects. To mainter -16 hoursof extrememagneticquiet. The data presented herein are from the DefenseMeteo- tain operational unambiguity,and because the zero-energy rologicalSatelliteProgram(DMSP) F7 polar-orbiting satel- boundaryhas theoreticalimportance its own right, we in lite, which was in a nearly circular 835-km-altitudeorbit do not stress here the connection with the plasmapause. with 98.7Ü inclination. F7 was Sun synchronous, lying in Boundary 2e, the poleward edge of the dEe/dX > 0 the 1030-2230 magnetic local time (MLT) plane. The region (1125:47 UT). It has long been known that lowSSJ4 particle detectorsonboard measuredelectrons and energy electronsin the plasma sheet reach closer to Earth ions from 32 eV to 30 keV in 19-point spectrawith 1-s than do higher-energy electrons[Vasyliunas,1968; Schield resolution[Hardy et al., 1984]. The satellitewas three-axis and Frank, 1970; Fairfield and Vinas, 1984]. The higher the stabilized,with the SSJ4 aperturealways pointed toward the energy of the electronsmeasured, farther from the zenith so that at aurorallatitudesonly particleswell within Earth they appearto have a cutoff (some exceptions exist, the lossconewere sampled. suchas when a dispersionless injectionoccurs).As a lowaltitudespacecraft movespolewardfrom boundaryl e, progressively higher-energy electrons are observed, so that 2. Structure of Nightside Precipitation: Physical dEe/clX 0 (where E· is the averageenergy of the elec> Significance of the Boundaries trons). As one reachesthe main plasmasheet,electrons of As in many situations, exampleis clearerthan words all energies are observed. Farther poleward the overall an alone;hence consider Plate 1, which presents DMSP F7 trend is for dE·/dX < 0 (the region of negative gradient a pass at 1123 UT on January3, 1985 during a recovery has a slope of smaller magnitudeand exhibitsmore flucphase.Plate 1 is presented a typical case;a discussion tuationsthan does the region of positive gradient),simply as of the effectsof substorm cycle is deferred until section 4. becausethe plasma sheet is progressivelycolder farther We considerin turn each of the most geophysically sig- from the Earth. The point where dEe/dX= 0 is one meanificant boundaries that can be identified in this picture. sure of the start of the main plasmasheet(or in the termiBecause operational definitions which are robustrequirean nology of FG85, the true start of the centralplasmasheet). Boundary 2i, the high-energy ion equatorward preattentionto detail that is tediousto many, section2 gives conceptuallyoriented definitions, with the fine print re- cipitation cutoff or precipitating energy flux maximum
served for section 3.

Boundary 1, the zero-energy convection boundary (1127:13 UT). Zero-energyelectronsand ions have no curvature or gradient drifts, hence they should share a common equatorward boundary, one which is determined purely by the electric and magnetic field configuration (that is, the boundarywhich resultsonly from a consideration of ExB drift effects).The DMSP low-energyion detector has an extremely large geometric factor, which makes it possibleto observesuchcoincidences despitethe comparatively low fluxes of ions at low energies. Newell and Meng [1988b] have reported that in the duskand midnight sectorsthe electron and ion zero-energycutoffs indeed coincide on 80% of the passes. To maintain operational unambiguity,we proposethat the zero-energy electron and ion boundaries separately be defined (denotedb le and b1i, respectively).Then, when the two boundaries indeed coincide to within 0.25Ü magnetic latitude, one may reasonably say that a zero-energy convection boundaryexists. Any model electric or magnetic field of the magnetotailspecifiesa unique positionfor this boundary (as a function of MLT); hence observation the boundary of providesa direct comparison betweentheory and reality. In some theoretical formulations the zero-energy convection boundaryis also the plasmapause location [Nishida, 1966], and, indeed, observationsof electron data support

(1126:00UT). This boundary alsothe isotropy is boundary


NEWELL

ET AL.: MORPHOLOGY

OF NIGHTSIDE

PRECIPITATION

10,739

F7

85/3
ELECIRONS

LOG JE·
LOG AVE. E
2

LOG E FLUX
ELEC 10 ION 8

I
!t

!

ra 3
.i

E

2

,
·

o
r·

2

! t-..'
.

I

t t.t

!I

,

,

2

'

lil,
'I
lb6
UT 11:23:30 MLAT 75.1 GLAT 76.5 GLONG 202.1 MLT 22:45 11:24:04 73.2 75.0 196.5 22:43 11:24:38 71.2 73.3 192.4 22:43

3a l's
1:25:12 69.2 71.6 188.8 22:42

!·i ,., " tl
!b2i
1 !:25:46 67.2 69.8 185.9 22:42 11:26:20 65.3 68.0 183.4 22:42

!i
11:26:54 63.2 66.1 181.1 22:41

I4
11:27:29 61.3 64.2 179.3 22:41

5-':·,.3
Jan 3

Plate 1. A typical precipitation patternobserved crossing nightsideauroraloval by DMSP in the F7 on January3, 1985 at 1123 UT. Plotted are differentialenergyfluxes in units of eV/cm2 s sr eV (main panels);averageenergyin eV (bottomplot), and integralenergyflux in eV/cm s sr. 2

(IB) of Sergeevet al. [1983]. It is probablythe best and energy ions means that they scatter off field lines with mostdirectproxy for the location the earthward of edgeof smaller radii of curvature than do the lower-energy ions; the currentsheet.Consider ionsin the energyrangefrom a hence the higher-energy ions maintain isotropy farther
Neither the tail currentsheetnor the precipitating highdeclininglatitude,apparently a resultof adiabatic as accel- energy ions have a sharply defined boundary.Operationeration as plasma convectsearthwardin the magnetotail ally we proposeto use the ion precipitatingenergy flux [Galperin et al., 1978]. This steadytemperature increase peak (integratedover the range 3-30 keV), which univerboundaryof the highterminates with a relativelysharpequatorward precipitation sally occursnear the equatorward as cutoff. However,in the high-altitude inner magnetosphere,energyion precipitation, the definitionof b2i. The geoions do not disappearat the L-shell value of the high- physicalsignificance the boundaryis that it represents of a energy ion precipitationboundary(e.g., Lui et al. (1987]). good approximationto the earthwardedge of the tail curInstead,the ions becometrappedand ceaseto precipitate rent sheet. Sergeevand Gvozdevsky [1995] have demonin measurablequantities.Poleward of the precipitation stratedthat the latitudeof this ion isotropyboundaryhas a boundary any particularenergy,the ions are highly iso- very high correlation(r- 0.9), with the magneticfield inat tropic [Bernstein al., 1974; Sergeevet al., 1993]. It has clination(degreeof stretching) et measured simultaneously at thus been suggested,and even successfullymodeled in the geomagnetic equator. A conceptuallyclosely related boundaryis the >30-40somedetail, that the ions maintaintheir isotropyby pitch angle scattering in the tail current sheet [Lyons and keV electron-stable trappingboundaryfor electrons, which Speiser, 1982]. The physicalmechanism quite simple: we term b2t. Although we cannotdirectly identify it in our is ions cannot maintain pitch angle while bending around DMSP database, it has long been considered useful. field lines that have a radius of curvaturecomparable to Thought in the 1960s to representthe open/closedfield the ion gyroradii [Sergeevet al., 1983]. This explanation line boundary,b2t is now generallyrecognizedas another
also accounts for the dispersion the high-energyion cut- measure of where field lines begin to be significantly in offs [Sergeevet al., 1993]. The larger gyroradiiof higher- stretched(see the discussion pp. 246-247, FG85). Beon
few keV to tens of keV (30 keV for DMSP). Ions in this energyrange increasein temperature and energyflux with
earthward.


10,740

NEWELL ET AL.: MORPHOLOGY OF NIGHTSIDE PRECIPITATION

significant only if it also significantly exceedsthe background counts obtained by averaging over several equatorward seconds.If jep rises to a value above 8.0 "i" for ions. Better results are obtained if the terms (or if jip reaches 6.5), a factorof only 1.6 jump is acceptjep(E1, E2) etc. are introduced,referring to the "partial" able in determiningthe zero-energyboundary.If jep > that the boundary has been electronenergy flux betweenE1 and E2. Strictly speaking, 8.25 (tip > 6.9), it is assumed all instrumentsmeasure only partial values of "integral" reached, even if no jump in the value of the fluxes is parameters between the upper and lower ranges of their measured. detector (some confusion in the literature exists from the Special cases:The energy range considered (El to E2) neglect of this point, as in the case of the cusp electron in the partials dependson whether photoelectrons are average energy). As a practical consideration here, such present and whether the spacecraftis charged to -28 V. partials improve greatly the identification of boundaries The former can be identifiedby a sharpdrop-offin elecwhile also making the resultssomewhatless dependent on tron fluxes above 68 eV at latitudes below the auroral the specific detector (SSJ4) that they were designedfor. zone, the latter by a sharp cutoff above the 32-eV ion The units usedbelow are eV for E and 1og·0 eV/cm s sr channel. In the absence of these effects the channels are 2 for jep and jip. For each boundary,we start by giving the set to the lowest available value, i.e., E1 = 32 and E2 = rule that works most of the time and follow with caveats 47. If the spacecraft chargedto -28 V, the ion channels is and more detail on handlingspecialcases. are set to E1 = 47 and E2 = 68. If photoelectrons are Boundaries b le,i (zero-energy). The algorithm moves present(rare on the nightside),the next available "clean" from lower latitudes to higher, comparingthe averageof channelsare 100 and 145 eV, respectively.Finally, isotip(E1, E2) and jep(E1, E2) (ordinarily E1 and E2 are the lated noise can sometimes causefalse positives,as by ratwo lowest channels)over the three previousspectrawith diation belt counts at 1118:30 UT in Plate 4. Thus a the three succeeding spectra. An increase jip by a factor "check ble" routine performsa double-check simply in by of 2 marks the onset of the zero-energyboundary,which examining the next several seconds.If, in the next few is separately determined for the two species. This jump is seconds the auroraloval is purportedly as entered,a droptegral parameters are denoted n, JE, and E, referring to density, energy flux, and average energy, respectively. These quantitiescan take the subscript "e" for electrons or

F7

13 ........................................................... f'12
LOG E FLUX ELEC I0 ION 8

84/334

!

I

i

m3
,

·

2

{o
o

2.
z3
3

!

5

3

b2i,b 1
UT 05:19:00 MLAT -64.5 GLAT -69.9 GLONG 228.2 MLT 22:57 05:19:34 -66.4 -71.7 225.3 23:00

b2e
05:20:08 -68.3 -73.4 221.6 23:02 05:20:42 -70.1 -75.0 217.3 23:05 05:21:16 -72.0 -76.6 212.0 23:10

'b3a·b ·5
05:21:50 -73.8 -78.0 205.2 23:14 05:22:24 -75.6 -79.3 197.4 23:21

b6
05:22:59 -77.6 -80.2 186.6 23:27

Nov 29

Plate 2. The nightsideauroral oval under conditionsof prolongedquiet (November 29, 1984 at
0519 UT).


NEWELL ET AL.' MORPHOLOGY OF NIGHTSIDE PRECIPITATION
off in fluxes is exhibited instead of a rise in fluxes, the identification of b le is inaccurate, and the search resumes

10,741

Specialcases: Sometimes "nose"eventsoccur,namely, the injectionof high-energy ion regionsslightly detached
from the rest of the auroral oval [Konradi et al., 1975; Sdnchez et al., 1993]. Such events are identified as local

toward increasing latitudes. Boundary b2e (plasma sheet start). This algorithmlocatesthe first point polewardof ble where dE/dX _
maximafollowedby a declineand subsequent to glorise bal energy flux maxima. The detached(or partially detached) maximum is discarded. In addition, in times of

prolongedquiet, b2i may occur equatorward bli (as in of Plate 2), but by definitionb2e mustalwayslie poleward of
ble.

Special cases: If the correct boundary b2e has been found, one expectsto have enteredthe main plasmasheet.

Boundaries b3a,b (most equatorward and poleward

is Hence Jee 11 (orif both < 11.5andE· < 1000 if < j·r· eV), electron accelerationevents).Each individualspectrum
then the next 30 s are examinedto see if a spectrumexists examinedfor evidenceof a monoenergetic peak. This can witha higher aswellasa je·larger at least (i.e., either be a singlechannelwith a differentialenergyflux 5 E· by 0.3 a factor of 2 difference in the energy flux). Double check- times larger than any other or a sharpdrop by at least a ing is done only underthis suspicious circumstance low factor of 10 below the electron differential energy flux of fluxes becauseit risks encounteringa point with high E· peak. Details of this algorithm,includingspecialcases, are andhigh simply j·r· because strong field-aligned accelera-presentedin a separatepaper [Newell et al., 1996]. The tion is encountered. most equatorward and polewardindividualspectra showing Boundary b2i (ion isotropy boundary). This boundary such monoenergetic peaks are flagged as boundariesb3a is definedby the precipitation flux maxima for ions 3 keV and b3b, withoutregardto otherboundary locations. and above.The high-energy ions behavelessvariably than Boundary b4s (structured/unstructured boundary). do the electrons,so fewer precautions are needed.Thus a The countsin the variouschannelsfor a given spectrum slidingaverageof jip(3 keV, 30 keV) over 2 s is compared are correlated with the corresponding counts in the five with jip for any 3 consecutive seconds over the next 10 s previousspectra,and the five resultingcorrelationcoeffifarther poleward to determinewhether the maximum has cientsare averaged (). By definitionb4s lies poleward of b2e and b2i. When the sum of seven consecutive been located.The smallestmaximumacceptable 10.5. is

83/354

IONS

LOG E FLUX ELEC I0 ION 8

,

I

i-tl I.
..

..· j

2

5 .....3 t

b5T·'b3b .....
UT 21:00:20 MLAT 73.1 GLAT 78.4 GLONG 66.3 MLT 02:02
21:00:54 72.2 77.0 59.4 01:38 21:01:28 71.2 75.5 53.4 01:15

b3ai
21:02:02 70.0 73.8 49.1 00:56 21:02:36 68.7 72. I 45.2 00:38
21:03:10 67.2 70.3 42.2 00:23

21:03:44 65.7 68.5 39.5 ig):09

21:04:!9 64.1 66.6 37.2 23:57

Dec 20

Plate3. The nightside precipitation pattern shortly afteran onset which followed period proa of
longedquiet (December20, 1983 at 2100 UT).


10,742

NEWELL ET AL.: MORPHOLOGY OF NIGHTSIDE PRECIPITATION

dropsbelow 4.0, the searchis halted. b4s is set to be the farthestpolewardspectrum within the final groupof 7 s,
which has > 0.60.

boundaryby the trappingboundary.Auroral arcs apparently invariably occur poleward of b2i (cf. Lyons et al.
[1988]).

Specialcase:If the energyfluxes resultonly in an aurora which is subvisual < 10.7 or 0.25 erg/cm s), the (Je 2 correlationcoefficient is suppressed (halved); hence lowflux but homogeneous regionssuch as polar rain are auto-

Boundaries 3a,b, the most equatorward and poleward electron acceleration events. In the literaturemany proxies for identifying the region of discreteaurorasexist. For example it appears that most electron acceleration events, and certainly those of high accelerating potential matically excluded. Boundaries bSe,bSi (poleward edge of main oval). values,occur on the stretched field lines that lie poleward These boundariesare computedseparately, but using the of the >40-keV electron-stable trappingboundary[Frank sameprocedure. average JEfor the previous s is and Ackerson,1971]. It is quite feasibleto examineeach An of 12 and determinewhetherit showssignsof compared with Je for the succeeding s. When a drop-off electronspectrum 12 potential.Thereforewe include of a factor of 4 is located,a provisional boundaryis a field-aligned accelerating b5 determined. Note that this algorithmemphasizes locatinga boundaries 3a and 3b, which, based on the examination of each individual spectrum,are the farthestequatorward and sharpgradientin the flux levels. A Specialcases: The next 30 s are double-checked s farthestpolewardsitesof electronacceleration. spectrum (35 for electrons) make surethe drop-offremainsbelow au- is identifiedas accelerated it has either a monoenergetic to if roral energyfluxes.If the (log) average hasnot dropped peak or a sharpcutoff abovethe spectral jE peak (more detail below about 9.7 for ions or 10.5 for electrons, the search is available in Newell et aI. [1996]). Although we have it continues the corresponding boundary. for b5 Sucha large opinionsabout the likely locationof theseboundaries, is search ahead is needed because of features like the best that they be identified separately from all other pre"doubleoval" [Elphinstone al., 1995] in which fluxes cipitationboundaries. et causeof the smallergyroradii of electrons, b2t usually lies Boundary 4s, the onset of spatial structure in eleca short distancepolewardof b2i. There is somereasonto tron precipitation (on a scale of >_5-10 kin) (1125:14). believe that the stable trapping boundaryapproximately The BPS/CPS distinction,no matter what misconceptions corresponds the WYAH75 boundary between CPS and becameassociated to with it, has persisted primarily because BPS; and, indeed, Weiss et aI. [1992] define this latter many nightsidecrossings seem to have a highly strucdo

'12
ELECTRONS if)NS

84/346

LOG JE"
8

LOG E AVE

LOG E FLUX ELEC I0 ION 8

u

21

t
g

'

I

2
3

Z;3

IIII 4II t111! , , bl b2etb2i
UT I. 1:17:30 MLAT -57.9 GLAT -45.8 GLONG 154.3 MLT 22:40 11:18:04 -59.9 -47.8 153.5 22:40

i

, Ii I

,
b4s Ja3b I
11:19:46 -66.0 -53.6 150.8 22:4 ! 11:20:20 ·8.1 -55.6 149.7 22:42

Iii It
11:21:20 -72.4 -59.5 147.2 22:42

5

3

b5 ?
11:20:54 -70.2 -57.5 148.6 22:42

11:18:38 -61.9 49.7 152.6 22:40

11:19:12 ·.0 -51.7 151.8 22;41

Dec 11

Plate 4. The nightside auroraloval duringconditions moderate of activity(December 1984 at 11,
1117 UT).


NEWELL ET AL.: MORPHOLOGY OF NIGHTSIDE PRECIPITATION

10,743

tured region as well as a relatively unstructured region of in nature.)Sometimes, especially quiet ti·nes,including in auroralelectronprecipitation. this structured/unstructured If quieting times following a substorm, subvisualdrizzle a distinctionrepresents something fundamental, there should can extendpolewardof the oval. The conceptual definition be a quantitativeway of making the distinctionwithin the of the poleward edge of the main auroral oval is that the precipitation data themselves,i.e., one that does not de- precipitatingfluxes drop by a factor of at least 4 over a pend on such additionalfactorsas the boundaryof the ra- shortdistanceto valuesbelow 3 x 10·ÜeV/cm2 s sr (elecdiation belts (b2t). trons)or 10·ÜeV/cm s sr (ions).We emphasize even 2 that To move from a qualitativedescription (structured) a for northwardinterplanetary to magneticfield (IMF) condiquantitativedescription,we investigatedthe behavior of tionssuchsharpdrop-offs still occurand separate main the the correlationcoefficientof individual spectrawith their auroraloval from the polar cap precipitation (which, to be neighbors. Figure 1 showsa plot of the runningaverageof sure,usuallyresembles structured the plasmasheetprecipithe correlationcoefficientof each spectrum Plate 1 with tation, although in not at the intensityof the main oval). the previousfive spectra.(Our algorithmsuppresses specBoundary 6, the poleward boundary of the subvisual tra with fluxes below 0.25 erg/cm s by halvingthe corre- drizzle (1123:53 UT). Poleward of the main auroral oval 2 lation coefficient.) Although even within the BPS region is a weak subvisual drizzle that differs from polar rain in most of the individual spectrado not show evidenceof severalways. The subvisual drizzle usuallyincludesweak field-aligned acceleration,Figure 1 shows that the entire ion as well as electronprecipitation; the drizzle is strucregion is indeed structuredin the sensethat each point tured, whereaspolar rain is comparatively homogeneous; correlates only poorly with its neighbors. We thereforein- and the typical electronenergies a bit higherthan norare troduceboundary4s, the structured precipitation boundary, mal for polar rain. Polar rain is intensepolewardof the defined by the point where the correlation coefficient dayside oval and declines gradually in intensity as it drops from the 0.95-0.60 range to below 0.60. Thus b4s movestoward the nightsideoval. The subvisual drizzle exmarks a fundamentalchangein the characterof the elec- tendspoleward from the nightsideoval and terminateseitron precipitation, from a spatiallyunstructured regionto a ther when fluxes drop to background levels or (less comhighly structured one. Section5 points out that some lim- monly) when a smooth,coherent electronpolar rain signaited fine-scale structuremay appear within the generally ture is encountered. unstructured precipitationregion; in this senseit is "unstructured" over spatialscales >5-10 km. 3. Operational Definitions of Nightside Boundaries 5e,i, the poleward boundaries of the Precipitation main auroral oval (1124:36 UT). The precipitating energy flux in the auroraloval typically dropsby about an order In this sectionwe give the detail requiredto operationof magnitudeover a short distance(usually <-0.2Ü). This ally define the various boundaries unambiguously. The indramaticdrop-off occursin both the electrons and ions, al- temporarily drop below oval levels only to rise again though not always in preciselythe same location for the clearly to oval values. two species. active times, polewardof this sharpdropIn Boundary b6 (the poleward boundary of the suboff, there is often only a narrow region of high-energy visual drizzle). This boundaryis definedby the point (-10-keV) but diffuse electrons at very low flux levels. where either polar rain is encountered (identifiedby the (This faint "overhang"of high-energyelectronsseemsto
be both unremarked in the literature and a common feature

presenceunstructured of electrons noions) Jee and or drops below 10.4and i drops jE below 9.6.
Specialcases: The drizzle is definedby weak structured fluxes with ions and electronsabove noise levels. By checking the computed average energy, one can infer whether noiseis significant: exampleif counts ranFor are
domly distributedacrossall channels,one would obtain E = 15 keV for DMSP. If Ee< 500 eV, even a flux as low as 10.0 is acceptable. lower average A energyimpliesthat a lowerminimumflux level is interpretable physical. as

Correlation with Neighboring Spectra
1.0 · ·

0.8

0.t5
0.4

4. Geomagnetic Activity and the Stagesof Nightside Precipitation

Occasional intervals prolonged of quietoccurin which manyhourselapsewith no apparent auroralactivity.Plate 2, a spectrogram from November 29, 1984 at 0519 UT, 0.0 showsthe precipitation observed during an extremeexampleof magnetic quiet.The lastprevious substorm ended 0 50 100 150 200 some16 hourspreviously, leastjudgedby the flat and at utsec(from 11:23:00) near-zero valuesfor A U andAL in the intervening interval. Figure 1. A plot of the averagecorrelation coefficientof In this unusual case the entire CPS as defined by levels--theregionof each electron spectrumwith its five predecessors the WYAH75 is beloweasilymeasurable for pass shown in Plate 1. The dashed line shows the hot but highly correlated electronprecipitation (between polewardboundary structured of precipitation, b4s. b2e and b4s) has virtuallydisappeared. Also goneis the
0.2


10,744

NEWELL ET AL.: MORPHOLOGY OF NIGHTSIDE PRECIPITATION

dEe/dX> 0 region (between le and b2e), along with the trons of severalkeV, even in individual spectrathat do not b downwardposimilarlydispersed ion precipitation soft (between li and contain the classic signs of an accelerating b
tential. In general, as reportedby WYAH75, only during quiet (or quieting) conditions doesthe structured region of precipitation (between our b4s and b5e) consist of comparatively soft plasma (kTe < -300 eV, kTi- 1 keV) (we have added informationon the ions to the description of WYAH75). Although we agree with the WYAH75 descriptionof this phenomenon, is not clear whether the it with electron acceleration events and contain a large original explanationis correct,that is, whetherthe disapBPS" reflectsthe energization amountof O+ [Sauvaud al., 1981; Bosqued al., 1986]. pearanceof the "deactivated et et The sharppolewardcutoffsto Once introduced into the magnetosphere, ions are ExB of that plasmapopulation. the that often appearduring convected earthward, and henceto lower latitudes. is ap- the electronand ion precipitation It activity are reasonably ascribed an open/closed to propriate that FG85 termed regionbetween the boundaries substorm 1 and 2 "the remnantlayer" sincetheseequatorward ions field line boundary;it may thereforebe that the ionosphere and electronsdo seem to be the remnantsof magnetic ac- is no longer magnetically connectedto the more distant tivity (including the effects of earthward convection plasma sheet population that originally suppliedthe soft precipitation. from the plasmasheet). Our final example is Plate 4, representing DMSP F7 a The extendedregion of soft electrons within the main auroral oval observable in Plate 2 corresponds what pass from December 11, 1984 at 1117 UT, representing to WYAH75 termed the "unactivatedBPS." The generalpro- moderatelyactive auroral conditions.(Note that the phefile of the plasmasheetpopulation more faithfullypre- nomenonobservedat 1118:32 UT is causedby MeV elecis servedin this instancethan in the more typical (lessquiet) trons from the radiationbelt penetrating detectors the and b2i). This latter coincidence betweenboundaries and 2 is 1

not highly unusual. The extended regionof dispersed cutoffs disappears after only a few hourswith no substorm activity [Newell and Meng, 1987; Sdnchez al., 1993]. et There is goodevidence that the ions observed between l b and b2 are originally injected from the ionosphereat higher latitude (say betweenb4s and b5) in association

situation: beyondabout69Ü magnetic latitudethe ions and electrons have a temperature about 1 keV and 200 eV, of respectively, valuesquite appropriate the plasmasheet to for say 10 Re and beyond(e.g., Baumjohann al. [1989]; et Christon et al. [1989]). Only a few acceleratedelectron spectra(in this case, toward the polewardedge of the
oval) are to be observed.

doesnot represent precipitation.) The dEe/dX> 0 region is reasonablywell developed,along with the associated low-energyion injection.The fairly sharppolewardcutoff
at 1120:58 UT is characteristic of active times and has

Generallythe first changenoticeable with the onsetof auroralactivityis the injection hot plasma(10 keV and of up for ions, several keV for electrons) into the equatorward (near-Earth) region of the auroral oval (Lasseneta!. [1977] studiedthe relationship suchinof jectedhot plasma auroral to phenomena seen all-sky as by cameras and ground-based magnetometers). Sometimes in the premidnight localtime sector ionscanbe observed hot priorto the observation hotelectrons, happened be of as to the casefor the next DMSP passafter Plate2.
To illustrate more of the featuresobservedshortly after

reasonably been interpreted representing open/closed as the field line boundary. In this case one can observe at the poleward edge of the auroral oval what is a fairly common feature of active times, yet seemsto be unmentioned in the literature. Notice that there is a faint highenergy (-1-15-keV) electron "overhang"present.Reports in the literature generally describethe polewardsubvisual "polar diffuse aurora" as owing to low-energyelectrons, which they often surely do. However, in the courseof creating automatedboundaryidentifications the nightside for oval, we found that this narrow and diffuse high-energy electronoverhang frequentduringactivetimes. is Figure 2 shows that the correlationcoefficient allows for a rigorous separationbetween the structuredand un-

onsetwe present differentinstance substorm a of onsetafter a period of prolongedquiet in Plate 3, showinga DMSP F7 passfrom December 1983 at 2100 UT (the 20, expansion phase).In this slightlypostmidnight it is case the injection the hot electrons the equatorward of in region
able between 2101:37 and 2102:50 UT, and, farther

structured portionsof the auroral oval. Prior to 1120:06 UT each spectrum correlated with its five preceding neigh-

bors on the level of 0.6-0.98, albeit with occasionaldips. Poleward of 1120:06 UT the correlationdroppeddramatically, to the range generallybelow 0.5, althoughwith ocof the auroral oval which is more dramatic. Plate 3 shows casionalspikesabovethat level. It remainsto discuss precipitation patternsunderquiettwo othernoteworthy features: auroral an bulgeis observ-

ing conditions following a substorm. Actually Plate 1 reppoleward, electrons now in what WYAH75 termed resentssuch a case. The extendeddEe/dX > 0 region is the are activityrecentlyoccurred, whereas the an "activated"state.The form that an auroralbulge caused proof that substorm in by substorm expansion takes in satelliteprecipitation ob- lack of energization the polewardregion of the strucservations is an extended interval of electron acceleration tured auroral oval occursonly under quieting conditions. (monoenergetic peak spectra) which the accelerating in po- In the latter respectthe behaviorof the auroraloval confollowsthe pattern first described WYAH75. by tential does not rise and fall in a "V"-shapedfashionbut sistently insteadshowsa relativelyconstant accelerating energy(cf. 5. Discussion Lopez et al [ 1991]). Further characterization of the nightside precipitaPoleward of the bulge feature (between2100:45 and 2101:35 UT) lies the highly structured precipitation charac- tion. The scheme outlined herein does not exhaust the useteristic of the activatedBPS of WYAH75. This region is ful characterizationof nightside precipitation. Some noare of regions, highly structured the sense in that the correlation between table omissions the existence flux depletion ion (VDISneighboring electronspectra low (typicallyin the 0.0- surge forms, and velocity-dispersed structures is 0.4 range).During activetimes this regioncontains elec- 2) at the polewardedge of the oval.


NEWELL ET AL.: MORPHOLOGY OF NIGHTSIDE PRECIPITATION

10,745

Correlation with Neighboring Spectra

0.8
0.4 0.2

which actually mean quite different things to different people. One minor point is that Lui et al. [1977] defined the diffuse auroraas a region exceeding1 kR in intensity,
whereas FG85 defined the diffuse aurora as a "subvisual"

region at the equatorward edge of the aurora.This latter definitionwas given partly because relativelyhomogeneous emissions difficult to resolvefrom ground-based are observations ("subvisual"in the senseof being "unresolvable"), and partly becausethe actual intensity of the diffuse aurora is in dispute.The disagreement stemsfrom differing
definitions of the term "diffuse aurora."

0

50

1O0

150

200

utsec(from 11:17:30)

Figure 2. The correlation coefficient of each spectrum with its neighboringpredecessors the pass shown in for
Plate 4. The dashed line illustrates b4s.

The

identification

of VDIS-2

events is of interest be-

causeit has been plausibly arguedthat such instances occur on field lines mapping to the high-altitude PSBL [Zeleyni et al., 1990], although observationsof VDIS events seem to be a sufficient but not a necessary condition for making such a mapping [Burke et al., 1994].
VDIS-2 structures are to be found between boundaries 5

and 6 whenthey are present. example the DMSP An in
data set can be seen in Plate 1 of Senior et al. [1994]. As

yet the criteria for identifying a structureas VDIS-2 are quite vague; it is not an easy thing to look at a spectrogram and infer that the ions exhibit a dispersion due to a velocity filter effect. We believe legitimateVDIS-2 signatures exist, but this matter needs careful consideration before code is introduced to automate their identification.

Flux depletionregionsare characterized deep drops by
in both electron and ion flux levels well within the auroral

oval. These are distinguished from the drop-off in ion flux associated with the potentialdropsthat accelerate electrons downward.An example of the latter phenomenon can be havecombined ground-based satellite and measurements to seenin Plate 3; an exampleof the former can be seenin show that a discretearc can appeardeep within an origiPlate 14 of Sdnchez et al. [1993]. It is believed that flux nally unstructured precipitation region(the satellite CPS). This difference in how the term diffuse aurora is dedepletion regions correspond to rarefactions in the magnetotail. fined explainswhy, according FG85, nearly all of the to Plate 3 also servesto illustratean auroral bulge associ- precipitating energyflux into the ionosphere occursin the ated with the substorm expansion phase(note that ions are oval of discreteforms, whereasto many satelliteresearchretardedfrom precipitatingin the bulge). The distinctive ers significant energyflux into the ionosphere deposited is structure such an expansion of bulge is a fairly constant in the diffuse portion of the auroral oval. We believe that both the Feldstein and Starkov [1967] downward-accelerating potentialover a large spatialscale (as opposed the rise and fall of inverted to V-shaped struc- definition of the oval of discrete forms and the satellitetures). The scale of bulge forms (up to severalhundred baseddefinitionof spatiallystructured/unstructured precipikilometers)is larger than typical of invertedV structures, tation are valid, but it is now clear that these definitions although latter can occasionally of comparable the be scale. only approximatelycoincide. Moreover, it is known that Becauseof their ease in identificationand importancein auroral arcs as observed from the ground are narrower substormphenomenology,these events can also be in- than can be resolved from most satellite observations. In cluded in the automated monitoring of the nightside fact the scale of auroral arcs is too small to reflect any magnetospheric structure[Borovsky,1993]. Becausethe magnetosphere. Discrete and diffuse aurora. "Discrete" and "diffuse" arcs tend to cluster into larger-scalegroupings,in most aurora are terms thought to be universallyunderstood but cases a good relationship probably exists between the

FG85, in keeping with Feldsteinand Starkov [1967], identify the "oval of discreteauroral forms" as beginning when the first auroral arc occurs and extending to the polewardedge of the main oval. All precipitation between b3a and b5 (or perhaps b2t and b5) is part of the "discrete auroraloval." This terminologyis perfectlyreasonable and widely used; however, one must then rememberthat what FG85 term the discreteauroral oval may includeportions of what many satelliteresearchers separate out as the diffuse aurora(includingall precipitation betweenb2e and b4s). Most individual spectrawithin this latter region do not exhibit signs of field-aligned electron acceleration or, indeed, any spatial structure on a scale 5-10 km or greater(i.e., each spectrum highly correlated is with its neighbors). Qualitatively similar results have been obtained by Elphinstoneet al. [1995] using simultaneous particle and imager observations. They report that, for the particular case they were studying,"the region of unstructured to 1 10 keV electronprecipitation overlapswith the main UV auroral oval." This implies that the diffuse aurora is intense and energetic,as originally introduced Lui et al. by [1973]. Elphinstoneet al. [1995] also agree that discrete auroral featurescan appearwell inside the "CPS" region as traditionallydefined. Likewise, an examinationof Plate 2 of Lopez et al. [1991] showsan invertedV structure (at 0302:20 UT) deep within the region they identify as diffuse aurora--andreasonably sinceit is within the dee/ so, d'^ > 0 region.Thereforeunderthe Feldstein and Starkov [1967] definition, the region of diffuse aurorafrom satellite observations in this caseactuallypart of the oval of is discreteforms. As a final example,Samsonet al. [1992]


10,746

NEWELL ET AL.: MORPHOLOGY

OF NIGHTSIDE PRECIPITATION

large-scaleelectron accelerationevents (5 to several hun-

dred kilometers) observed satelliteand the very smallby scale discrete arcs (tens or hundredsof meters) observed from the ground. However,to promote precision and avoidapparent contradictions, suggest we limiting the terminology used according to the measurementsmade. The term "auroral arcs"is bestreserved useby ground-based for researchers, or at least by thosewith the resolution capabilityto identify -100-m structures (and to determinethat the structure is extended longitude). in The satellitecommunity can instead employ the terms "electron acceleration events,"
"structured aurora," and "unstructured aurora." The first

keV ion-precipitating energy-flux maximum, whichalways occursnear the equatorward edge of the oval, just before the high-energy population ion becomes trapped. The ion isotropy poleward b2i results of from scattering the curin rent sheetwherethe gyroradius comparable the magis to netic field line curvature; hence,b2i is a proxy for the current sheetearthward edge.A closely relatedboundary is
the outerboundaryof the Van Allen radiationbelt, defined as the stable trappingboundaryof >30-40-keV electrons

(b2t). This also is physically determined scattering by in
the magnetotailcurrentsheet.

term is quantifiableby the presenceof a monoenergetic peak or at least a sharpcutoff abovethe electronspectral peak. The structure lack thereofcan be determined or by the correlationcoefficientof individualspectrawith their neighbors.Regions lacking accelerated electronsand for which eachspectrum closelyresembles neighbors its (correlation coefficient0.6-0.95) are thus the regionsof unstructured aurora. Note that not all structured aurora need

show signsof electronacceleration.

Finally, the term "diffuse" aurora was introduced by Lui and Anger [1973] in the contextof satelliteimagery. Since the spatialresolution suchimageryis generally of comparable (or somewhat to below) satelliteparticledata, it is likely that the diffuseauroragenerally corresponds to the region of unstructured particleprecipitation. However, this is not universallythe case,because is possibleto it have accelerated electrons with a fairly constant potential drop in auroralbulge featuresassociated with the substorm
expansion phase.Thus Bythrowand Potemra [1987] identify a large part of a DMSP opticalauroralimageas diffuse aurora,althoughan examination the particledata of showsthat the electrons evinceclear monoenergetic peaks throughoutthe region of bright emissions. Becauseonly
spatial structure and not accelerationfeatures can be inferred from the imager, it is occasionally possiblefor a diffuseaurorato be "structured" the spectral not the in but spatial sense. This phenomenonis comparatively rare,

Boundaries b3a,b3b. The most equatorward and poleward electron spectra which show signs of fieldaligned acceleration through a potential drop (monoenergetic peaks)are termedboundaries and b3b. b3a Boundary b4s. Electronprecipitation near and sometimespoleward boundary oftenlacksspatial of 2 structure on a scale length of 5-10 km or larger, in the sensethat eachspectrum correlates highly with its neighbors (correlation coefficients the range0.6-0.95). Farther in poleward, the electronsare more highly structured, with correlation coefficients betweenneighboring spectra generallyin the 0.0-0.4 range. The dividingpoint betweenstructured and unstructured electron precipitation boundary is b4s. BoundariesbSe,bSi.The contiguous oval generally has a sharppolewardcutoff, with a drop in fluxes by a factor of at least 4 over a -0.2 Ü latitudinalrange.Althoughthe cutoff is sharperfor southward than northwardIMF, even in the latter casethe drizzlepoleward the oval is typiof cally an orderof magnitude lessthan anything observed in the main oval. Thus a clearpoleward boundary the oval to
exists, even under northward IMF conditions.

Boundary b6. Under active conditions,boundary5 usuallyrepresents polewardboundary precipitation, the of exceptfor a very narrow "overhang" regionof weak electron precipitation the -10-keV range. Often, especially in under quiet or quieting conditions,there is a region of low-energy electronand ion precipitation low flux levels at

(1-3 orders of magnitude less intense than in the main oval), which is highly structured the senseof having (in however. neighboring spectrathat are poorly correlated). The precipitation between b5 and b6 is termed the subvisual drizzle poleward of the oval. Boundaryb6 is defined as 6. Summary and Conclusions the point wherefluxes eitherdropto levelsnot easilymeais In this paper we have proposed revisedphenomenol- surableor drop until a polar rain signature encountered. a The terms diffuse aurora, discrete aurora, auroral arcs, ogy of nightside precipitation including boundariesineventsare being usedin the littendedto facilitate the abstraction geophysically of signifi- and electronacceleration cant information.We introduced specificquantitative algo- erature in ways that, under some conditions,can lead to The ground-based "auroraloval of rithms to identify these boundariesin order to minimize apparentcontradictions. forms" sometimes includes part of the regionconoperational ambiguity. Quantitative details are given in discrete section 3, and justification (including appropriaterefer- sidered to be diffuse aurora from satellite observations. We ences)is given in sections and 4. A compactsummary recommend that the term auroral arcs be limited to use 2
follows.

when

instrumentation

of resolution

-100

m or better

is

available. Particle observations can identify an "unstructured" aurora, a term we have quantified as a region curvature or gradient drifts; hence, their earthwardextent whereinthe correlation coefficient eachindividualspecof is determined solelyby ExB drifts. The zero-energy elec- trum with its neighborslies in the 0.6-0.95 range. "Structron boundary often coincides with the plasmapause. tured" particle observations then those with a correlaare Boundariesb2e,b2i,b2t.In the equatorward portionof tion coefficient below 0.6 (generally the range0.0-0.4). in the auroralprecipitation, dEe/dX> 0; in the poleward por- Such structure can be causedby field-alignedelectronaction, dEe/dX< 0. The point wheredEe/dX= 0 is termed celeration,but not all structured spectrashow such signs. b2e. One interpretation that b2e is the start of the main Finally, the term diffuseauroraoriginallyreferredto satelis plasmasheet.A boundaryoften locatednearby is the >3- lite imageryin which spatialstructure was not identifiable.

Boundaries ble,bli. The geophysical significance of theseboundaries that zero-energy is particles experience no


NEWELL ET AL.: MORPHOLOGY

OF NIGHTSIDE PRECIPITATION

10,747

It is reasonable restorethis originaldefinition,taking Feldstein,Y. I., and Y. I. Galperin,The auroralluminosity to structure the high-latitude in upperatmosphere: dynamIts note that a lack of spatialstructure an imageusually, in ics and relationship to the large-scale structureof the but not always,impliesa lack of spectral structure the in Earth's magnetosphere, Geophys., 217, 1985. Rev. 23, corresponding precipitation.
Acknowledgments. We have benefited from discussions 15, 209, 1967. with L. A. Weiss,A. T. Y. Lui, K. Makita, and especially V. Feldstein, Y. I., and G. V. Starkov, The auroral oval and the Sergeev. Work at APL was supportedby AFOSR grant boundary of closed field lines of geomagneticfield, F49620-92-J-0196. I. Feldsteinacknowledges grant Y. ISF Planet. SpaceSci., 18, 501, 1970. M6P300, and Y. I. Galperinacknowledges RFFR grant94-02- Frank, L. A., Plasmain the Earth'spolar magnetosphere, J. 04299a. The DMSP SSJ4 instrument was designed and built Geophys. Res., 76, 5202, 1971. by D. Hardy of the PhillipsLaboratory. Frank, L. A., and K. L. Ackerson, Observations charged of The editor thanksJ. D. Winninghamand E. M. Basinskaparticle precipitationinto the auroral zone, J. Geophys. Lewin for their assistance evaluating paper. in this Res., 76, 3612, 1971.

Feldstein,Y. I., and G. V. Starkov,Dynamicsof auroralbelt and polar geomagnetic disturbances, Planet. Space Sci.,

Galperin,Y. I., and Y. I. Feldstein, Auroral luminosity and
References

Galperin, Y. I., V. N. Ponomarev, Y. N. Ponomarev,and Res., 94, 6597, 1989. A.G. Zosimoval,Plasmaconvection the eveningsector in Bernstein, W., B. Hultqvist,and H. Borg, Some implications of the magnetosphere the nature of the plasmapause and of low altitude observations isotropicprecipitation of of (in Russian), Cosmic Res., 13, 669, 1975. ring current protons beyond the plasmapause, Planet. Galperin, Y. I., N. V. Jorjio, R. A. Kovrazhkin,F. Cambou, SpaceSci., 22, 767, 1974. J. A. Sauvaud,and J. Crasnier,On the origin of auroral Borovsky, E., Auroralarc thicknesses predicted variJ. as by protonsat the day-side auroral oval, Ann. Geophys.,32, ous theories, Geophys. J. Res., 98, 6101, 1993. 117, 1976. Bosqued, M., J. A. Sauvaud, D. Delcourt,Precipitation J. and of superthermal ionospheric ions accelerated the conju- Galperin, Y. I., J. Crasnier, Y. V. Lissakov, L. M. in Nikolaenko, V. M. Sinitsin, J. A. Sauvaud, and V. L. gate hemisphere, Geophys. J. Res., 91, 7006, 1986. Khalipov,Diffuse auroralzone, 1, Model of the equatorial Burch, J. L., Low-energy electronfluxes at latitudesabove boundaryof diffuse auroral electronprecipitation zone in the auroralzone,J. Geophys. Res., 73, 3585, 1968. the evening and near midnight sectors(in Russian),CosBurke, W. J., J. S. Machuzak, N. C. Maynard, E. M.
Basinska, G. M. Erickson, R. A. Hoffman, J. A. Slavin,
mic Res., 15, 421, 1977.

Baumjohann, G. Paschmann, C. A. Cattell, Average W., and plasmaproperties the centralplasmasheet,J. Geophys. in

its relationshipto magnetospheric plasmadomains,in Auroral Physics,edited by C.-I. Meng, M. J. Rycroft, and L. A. Frank, p. 207, CambridgeUniv. Press,New York,
1991.

and W. B. Hanson,Auroral ionospheric signatures the Galperin, Y. I., V. A. Gladyshev, N. V. Jorjio, R. A. of Kovrazhkin, V. M. Sinitsin, F. Cambou, J. A. Sauvaud, plasma sheet boundary layer in the evening sector, J. and J. Crasnier,Adiabaticacceleration inducedby convecGeophys.Res., 99, 2489, 1994. tion in the plasmasheet,J. Geophys. Res.,83, 2567, 1978. Bythrow, P. F., and T. A. Potemra, Birkeland currentsand energetic particles associated with optical auroral signa- Hardy, D. A., L. K. Schmitt, M. S. Gussenhoven, F. J. Marshall, H. C. Yeh, T. L. Shumaker, A. Hube, and J. turesof a westwardtravelingsurge,J. Geophys. Res., 92, Pantazis,Precipitatingelectron and ion detectors(SSJ/4) 8691, 1987. for the block 5D/flights 6-10 DMSP satellites: Calibration Cambou, F., and Y. I. Galperin, Overall results from the and data presentation, Rep. AFGL-TR-84-0317, Air Force ARCAD experimentaboardthe satelliteAUREOLE, in InGeophys.Lab., HanscomAir Force Base,Mass., 1984. ternational Symposium on Solar-Terrestrial Physics, Heikkila, W. J., and J. D. Winningham, Penetration of p. 396, Sao Paulo, Brazil, 1974. magnetosheath plasmato low altitudesthroughthe dayside Cambou, F., and Y. I. Galperin, Main resultsof the joint magnetospheric cusps,J. Geophys. Res., 76, 883, 1971. French-Sovietspaceproject ARCAD-1 and ARCAD-2 for magnetospheric,auroral and ionospheric physics, Ann. Horwitz, J. L., S. Menteer, J. Turnley, J. L. Burch, J. D. Winningham, R. Chappell,J. D. Craven,L. A. Frank, C. Geophys., 87, 1982. 38, and D. W. Slater, Plasmaboundaries the inner magnetoin Christon, S. P., D. J. Williams, D. G. Mitchell, L. A. Frank, sphere, Geophys. J. Res., 91, 8861, 1986. and C. Y. Huang, Spectralcharacteristics plasmasheet of R. ion and electronpopulations during undisturbed geomag- Johnson, G., R. D. Sharp,M. F. Shea,and G. B. Shook, Satellite observations two distinctdaysidezonesof auof netic conditions, Geophys. J. Res., 94, 13,409, 1989. roral electronprecipitation (abstract), Eos Trans.AGU, 47, Deehr, C. S., J. D. Winningham, F. Yasuhara, and S.-I.
Akasofu, Simultaneous observations of discrete and diffuse 64, 1966.

Jorjio, N. V., J. Crasnier,J. A. Sauvaud,and V. M. Sinitsin, Auroral diffuse zone, III, Comparison betweenthe equatorial borderof the diffuse auroralzone and the plasmapause Eastman, T. E., L. A. Frank, W. K. Peterson, and W. positionduringFebruary13 and 17, 1972 magnetic storms Lennartsson, plasmasheetboundary The layer, J. Geophys. (in Russian), CosmicRes., 16, 937, 1978. Res., 89, 1553, 1984. Eather, R. H., Latitudinaldistributions auroraland airglow Jorjio, N. V., R. A. Kovrazhkin, M. M. Mogilevsky, J. M. of Bosqued,H. Reme, J. A. Sauvaud,C. Beghin, and J. L. emissions: The "soft" auroralzone, J. Geophys. Res., 74, 153, 1969. Rauch, Detection of suprathermal ionospheric O+ ions inside the plasmasphere, SpaceRes., 141, 1985. Adv. Elphinstone, D., et al., The doubleoval UV auroraldistriR. bution, 1, Implicationsfor the mappingof auroralarcs,J. Konradi, A., C. L. Semar, and T. A. Fritz, Substorm-injected Geophys. Res., 100, 12,075, 1995. protonsand electronsand the injectionboundarymodel, J. Geophys.Res., 80, 3055, 1975. Fairfield, D. H., and A. F. Vinas, The inner edge of the plasmasheet and the diffuse aurora,J. Geophys. Res., 89, Lassen,K., J. R. Sharber,and J. D. Winningham,The devel841, 1984. opment of auroral and geomagnetic substorm activity after aurorasby the Isis 2 satellite and airborneinstruments, J. Geophys. Res., 81, 5527, 1976.


10,748

NEWELL ET AL.: MORPHOLOGY OF NIGHTSIDE PRECIPITATION

a southwardturning of the interplanetary magneticfield Sauvaud, J. A., Y. I. Galperin, V. A. Gladyshev, A. K. following several hours of magneticcalm, J. Geophys. Kuzmin, T. M. Muliarchick, and J. Crasnier, Spatial Res., 82, 5031, 1977. inhomogeneity magnetosheath of protonprecipitation along Lopez, R. E., H. E. Spence,and C.-I. Meng, DMSP F7 obthe dayside cusp from the ARCAD experiment, J. Res., 85, 5105, 1980. servations a sub·torm of field-aligned current, Geophys. Geophys. J.
Res., 96, 19,409, 1991.

Sauvaud, J. A., J. Crasnier, K. Moula, R. A. Kovrazhkin, and

Lui, A. T. Y., and C. D. Anger, A uniform belt of diffuseauN. V. Jorjio, Morning sector ion precipitationfollowing roral emission seen by the Isis-2 scanning photometer, substorm injections, Geophys. J. Res.,86, 3430, 1981. Planet. Space$ci., 21, 809, 1973. Sauvaud, J. A., J. Crasnier, Y. I. Galperin, and Y. I. Lui, A. T. Y., D. Venkatesan,C. D. Anger, S.-I. Akasofu, Feldstein, statistical A studyof the dynamics the equaof W. J. Heikkila, J. D. Winningham,and J. R. Burrows,Sitorial boundaryof the diffuse aurorain the pre-midnight sector,Geophys. Res. Lett., 10, 749, 1983. multaneous observations particle precipitations of and auSchield, M. A., and L. A. Frank, Electron observations beroral emissionsby the ISIS 2 satellite in the 19-24 MLT tween the inner edge of the plasma sheet and the sector,J. Geophys. Res., 82, 2210, 1977. plasmasphere, Geophys. J. Res., 75, 5401, 1970. Lui, A. T. Y., R. W. McEntire, and S. M. Krimigis, Evolution of the ring currentduringtwo geomagnetic storms, J. Senior, C., D.C. Delcourt, J. C. Cerisier, C. Hanuise, J.P. Villain, R. G. Greenwald, P. T. Newell, and F. J. Rich, Geophys. Res., 92, 7459, 1987. Lyons, L. R., and T. W. Speiser,Evidence for current sheet Correlated observations the boundary of between polarcap and nightside auroral zone by HF radars and the DMSP acceleration the geomagnetic in tail, J. Geophys. Res., 87, 2276, 1982. satellite,Geophys. Res. Lett., 21, 221, 1994. Lyons, L. R., J. F. Fennell, and A. L. Vampola, A general Sergeev, A., and B. B. Gvozdevsky, V. MT-index, A possible association betweendiscreteaurorasand ion precipitation new index to characterize magnetic configuration of from the tail, J. Geophys. Res., 93, 12,932, 1988. magnetotail, Ann. Geophys., 13, 1093, 1995. Mauk, B. H., and C.-I. Meng, Dynamical injections as the Sergeev, V. A., E. M. Sazhina, N. A. Tsyganenko,J. A. Lundblad, and F. Soraas, Pitch-angle scatteringof enersource of near geostationaryquiet time particle spatial boundaries, Geophys. J. Res., 88, 100,011, 1983. getic protonsin the magnetotail currentsheetas the domiNewell, P. T., and C.-I. Meng, Energy dependenceof the nant sourceof their isotropicprecipitationinto the nightequatorwardcutoffs in auroral electronand ion precipitaside ionosphere, Planet. Space$ci., 31, 1147, 1983. tion, J. Geophys. Res., 92, 7519, 1987. Sergeev, A., M. Malkov, and K. Mursula,Testingthe isoV. Newell, P. T., and C.-I. Meng, The cusp and the cleft/boundtropicboundary algorithmmethodto evaluate magnetic the ary layer: Low altitude identification and statisticallocal field configuration the tail, J. Geophys. in Res., 98, 7609, 1993. time variation,J. Geophys. Res., 93, 14,549, 1988a. Newell, P. T., and C.-I. Meng, Categorization dispersion Vasyliunas,V. M., A survey of low-energyelectrons the of in curves in the equatorward edge of the diffuse aurora, evening sector of the magnetospherewith OGO 1 and Planet. SpaceSci., 36, 1031, 1988b. OGO 3, J. Geophys. Res., 73, 2839, 1968. Newell, P. T., W. J. Burke, C.-I. Meng, E. R. Sfinchez,and Weiss, L. A., P. H. Reiff, R. V. Hilmer, J. D. Winningham, M. E. Greenspan,Identification of the plasma mantle at and G. Lu, Mapping the auroraloval into the magnetotail low altitude,J. Geophys. Res., 96, 35, 1991a. using Dynamics Explorer plasma data, J. Geomagn. Geoelectr., 44, 1121, 1992. Newell, P. T., W. J. Burke, E. R. Sfinchez,C.-I. Meng, M. E. Greenspan,and C. R. Clauer, The low-altitude boundary Winningham, J. D., F. Yasuhara, S.-I. Akasofu, and W. J. Heikkila, The latitudinal morphologyof 10-eV to 10-keV layer and the boundary plasma sheet at low altitude: electron fluxes during magnetically quiet and disturbed Prenoon precipitation regions and convection reversal times in the 2100-0300 MLT sector,J. Geophys. Res., 80, boundaries, Geophys. J. Res., 96, 21,013, 199lb. 3148, 1975. Newell, P. T., S. Wing, C.-I. Meng, and V. Sigillito, The auroral oval position,structure, and intensityof precipitation Zelenyi, L., R. Kovrazkhin, and J. M. Bosqued,Velocityfrom 1984 onward: An automated online data base, J. dispersed ion beamsin the nightsideauroralzone: Aureole m3 observations, Geophys. J. Res., 95, 12,119, 1990. Geophys. Res., 96, 5877, 1991c. Newell, P. T., K. M. Lyons, and C.-I. Meng, A large survey of electron acceleration events,J. Geophys.Res., 101, in press, 1996. Y. I. Feldstein,Instituteof TerrestrialMagnetism, Ionosphere Nishida, A., Formation of plasmapause, magnetospheric and Radiowave Propagation, Troitsk, Russia. (e-mail: or plasma knee, by the combinedaction of magnetosphere gromova charley.izmiyan.rssi. @ ru) convectionand plasma escapefrom the tail, J. Geophys. Y. I. Galperin,Instituteof SpaceResearch the Academy of of
Res., 71, 5669, 1966.

Rostoker, G., and T. Eastman,A boundarylayer model for magnetospheric substorms, Geophys.Res., 92, 12,187, J.
1987.

Sciences of Russia, Moscow, Russia. (e-mail: YGALPERIN @romance.iki.rssi.ru) C.-I. Meng and P. T. Newell, JohnsHopkinsUniversity,Ap-

Samson, C., L. R. Lyons, P. T. Newell, F. Creutzberg, J. and
B. Xu, Proton aurora and substorm intensifications,

pliedPhysics Laboratory, Johns Hopkins Road,Laurel,MD 207236099. (e-mail:patrick.newell@jhuapl.edu)

Geophys. Res. Lett., 19, 2167, 1992. Sfinchez, R., B. H. Mauk, P. T. Nev;,ell,and C.-I. Meng, E.
Low-altitude observations of the evolution of substorm in-

jection boundaries, Geophys. J. Res., 98, 5815, 1993.

(Received April 27, 1995;revisedSeptember 1995; 14, accepted November15, 1995.)