Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.izmiran.rssi.ru/magnetism/Feldstein_2012/PDF/378.pdf Äàòà èçìåíåíèÿ: Wed Mar 11 16:41:28 2009 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 21:12:01 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: jet

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 102,NO. A7, PAGES 14,223-14,235, JULY 1, 1997

Auroral electrojets during geomagnetic storms
Y. I. Feldstein Institute of TerrestrialMagnetism,Ionosphere, and Radio Wave Propagation,Troitsk, Moscow
Region

A. GraIe
GeoResearch Genter, Potsdam, Germany

L. I. Gromovaand V. A. Popov
Institute of Terrestrial Magnetism,Ionosphere, and Radio Wave Propagation,Troitsk, Moscow
Region

Abstract.

On the basisof digital magnetometers from the Internationalmonitor

auroral geomagnetic effects (IMAGE) andEuropean incoherent scatter (EISCAT)
meridionalchainsin Scandinavia dynamicsof the eastwardand westwardelectrojets duringthe main phaseof magneticstormsare considered. the intensemagnetic For storm on May 10-11, 1992, with Dst = -300 nT, magnetograms subauroral of and midlatitudinal stationsLeningrad, Borok, and Moscowwere examined. It is found that the eastwardelectrojet center during the storm main phaseshifts

equatorward IDstl increases. as The electrojetcenteris locatedat the corrected geomagnetic latitude /b ,,· 59Ü-60 when Dst ,,· -100 nT and at /b ,,· 54o-55 Ü Ü
when Dst ,,· -300 nT. Data from meridional chainsof magnetometerssupport earlierresultspertainingto the relationship betweenthe westward electrojetcenter position and the ring current intensity for intervalsbetweensubstorms.During substormsexpansivephasesthe westwardelectrojet expandspoleward covering auroral latitudes/b ,,· 65Ü . The electrojetslocation during the storm main phase and their dynamicsin connection with substorms allow for interpretationsof effects described the literature: the AE indicessaturationduring the main phaseof in magneticstorms;approximately equal valuesof A U and Aoe indicesduring the storm initial phase and Aoe >> A U during the storm main phase.

Introduction

(2) differences the latitudes the observatories; in of and Kamide'andAkasofu [1983] andAkasofa al. [1983] et

istic featureof geomagnetic disturbances field [Chapman and Barrels, 1940]. The electrojets reachtheir

(3) limited of The auroral electrojet intensificationis a character- locations. latitudinalcoverage the AE observatories

jet maximum intensity at auroral latitudes and the auro- investigatedthe accuracyof the auroral electro indices.It wasshownthat duringrelativelyquiet periods, ral electrojet indices (AE , A U, Aoe) wereintroduced standard AE stations are not able to monitor properfor descriptionof the electrojetsby Davis and Sugiuly the activity of the aurorM electrojets. The ratio AE

as measures the auroral electrojet intensity and of of
The standard auroral indices are calculated on the basis

(12)/AE (71)(here AE (12)is AE ·dex determined using12 auror· observatories AE (71) is the same magnetospheric activity duringsubstorms storms. index determined on the basisanddata from 71 obserand of
vatories) sharplydecreases AE (12) < 400 nT. It for
arisesout of the polewardshift of the electro during jet quiet interv·s that causesthe standard observatories to escapefrom the ·one directly influencedby the electrojets. The resultsof the· investigationwas basedon data from s· meridional ch·n stations for March 17, 18, and 19, 1978, encompmsing both relatively quiet
interv·s and interv· with intense substorms.

ra [1966].Since then,they havebeenextensively used

of data from 12 magneticobservatories constitutinga longitudinal chainovercorrected geomagnetic latitudes between63Ü and 70Ü [Allen, 1970; Allen et al., 1976; Kamei andMaeda,1981]. The variationsof the electrojetindiceshave limited accuracy several for reasons: inhomogeneous (1) distribution of the AE observatories versusthe longitude;
Copyright 1997by theAmerican Geophysical Union.
Papernumber 97JA00577.
0148-0227/97/97JA-00577509.00

S·ce the auror· ov· expandsequatorwardand contracts poleward in msociation with the interplanetary magnetic field variations and substormsactivity, one

might expecta sharpdecrease AE (12) duringinin
14,223


14,224

FELDSTEIN

ET AL.: AURORAL ELECTROJETS

DYNAMICS

tensedisturbances well. Kamide andAkasofu as [1974] Table 1. Coodinatesof the Magnetic Observatories
analyzedthe latitudinal dependence the westwardauof roral electrojet profile acrossthe Alaska meridian chain of observatories, finding that during substorms with intensityas high as ~1000 nT, the maximumdeviationof the horizontalcomponentfrom the quiet level is located
at the auroral zone at ú ~ 65 Ü.

Geographic
Coordinates

Geomagnetic
Coordinates

Observatory
Soroya
Keva Masi

Code
SOR
KEV MAS

Lat
70.54N
69.76 69.46

Long
22.22E
27.01 23.70

Lat
67.21
66.15 66.04

Long
106.91
109.93 107.10

Akasofu [1981] studied Dst-AE relationship the for
a few intensemagnetic storms and found that at the

moderate stormlevel AE and IDstl growtogether in

Kilpisjarvi

KIL
MUO PEL

69.02
68.02 66.90

20.79
23.53 24.08

65.80
64.60 63.43

104.45
105.82 105.52

a practically linear manner. However, for more intense Muonio storms, he found that AE tends to saturate at a level Pello of ~ 1000 nT. It was suggested that as a storm devel- Oulujarvi Hankasalmi ops, the division of energyentering the magnetosphere Nurmijarvi between the ring current and the auroral ionosphere

OUL
HAN

64.52
62.30

27.23
26.65

60.86
58.60

106.60
105.02

changes.At the beginning a storm,when IDst[ is of
small, there is proportionality between the energy injection to the ring current and the auroral ionosphere. During the main phaseof intensemagneticstorms,most of the energygoesto the ring current. Similar AE-Dst relationshipas a function of the storm intensity is pre-

Leningrad

NUR LNN
BOX MOS

60.50 59.95
58.02 55.50

24.65 30.70
39.00 37.30

56.79 55.79
53.60 50.99

102.51 107.78
114.78 111.97

Borok Moscow

sented Gonzalez al. [1994]. by et Weigher al. [1990] et studied AE-B, relationship, Moderate Magnetic Storm on April 1993 the where B, is north-south component of the interplaneFigure presents 1 variations thenorthern comof (X) tary magneticfield. Saturation of AE takesplacewhen ponentof the geomagnetic alongthe IMAGE chain field B, is southward and intense. Such B, valuesare accom- during the moderatemagneticstorm with an initial
panied,as a rule, by the intensering current generation. phase onset that occurredbetween 0900 and 1000 UT

Auroral Electrojet Dynamics During a

txssxland Weimer et al. [1990]consid- on April 4, 1993. The storm main phase,connected with sharpdecreases the geomagnetic of field horizonamong others the shift of the auroral electro jets to- tal component low- latitudeobservatories, withat falls ward the equator during the main phase of magnetic in the interval1430UT on April 4 to 0700UT on April storms. However, this possibilitywas rejectedin fa- 5, 1993.Then,a slow recovery thequietlevelbegins, to vor of either energyredistributionor the nonlinearre- lasting several for days.The IMAGE chain during the sponse the magnetosphere-ionoSphere to the stormmain phasewasinitially locatedin the longituof system magnetospheric disturbances.Accordingto Weimer et dinalsector the eastward of electrojet (evening hours) al. [1990],the westward electrojetcenternevershifts andthenin thelongitudinal sector the westward of elecequatorward Sitka (·I, = 60Ü). of trojet (nightandearlymorning hours).'Vertical dotted Surnaruk al. [1989], et Feldsiein [!992],and Feld- lines mark latitudinal cross sections. The first two cross steinet al. [1994] have suggested theapparent that sat- sections andN2) characterize eastward (N1 the electroered various causesof the saturation effect, including

uration or evendecrease AE during the main phase of of intensestorms can be due to the auroral electrojets movingconsiderably lowerlatitudessothat the AE stations do not correctly monitor the substormevolution. For moreprecise calculations the AE indicesduring of magneticstorms,it is necessary use data from magto
netic observatories at subauroral latitudes where the au-

jet at the auroral ·one latitudes. The cross sections

N3 and N4 describethe geomagnetic field distribution relativeto the Hataug discontinuity, whereat latitudes equatorward the discontinuity of AX > 0 and poleward of the discontinuity AX < 0. Data in Figure 1 showa

clear shiftof theeastward directed current (AX > 0) to

the lower latitudes as the magneticstorm main phase roral electrojetsshift during magneticstorms. Below, develops the interval 1400-1800UT on April 4, 1993. in investigationof the eastwardand westwardelectrojets Besidestemporal UT effectsin the eastward electrodynamics continued is usingdata from the digital mag- jet locationscausedby the magnetospheric ring curnetic stationsmeridionalchain along the Scandinavian rent and the tail current development,LT effectsare peninsula (ElSCAT and later IMAGE) supplementedalsopossible from the asymmetryof the inner magneby analogue magnetograms Leningrad,Borok, and tosphereversusthe local time. Therefore the latitudinal of Moscow during an intensemagneticstorm. TheseRus- cross sectionN9 on April 5, 1993, was selected during sianstationsare locatednear the samemagnetic merid- the recovery phaseat the sameUT (1430 UT) as the ian. The aim of this paperis documentation the elec- crosssectionN1 during the storm initial phase. These of trojet dynamicsduring magneticstorms. The number difference betweentwo latitudinal cross section,asfar as of stationsequatorwardof the auroral oval along this location of the eastwardelectrojetis concerned, should meridianis uniqueat the present time. Table 1 presents be caused variations the intensityof the ring curby in the coordinatesof all stations usedin the analysis. rent and the tail current during the storm initial and


FELDSTEIN ET AL.' AURORAL ELECTROJETS DYNAMICS

14,225

06 apr' 1993
i I I

05 apr 1993

t

I

!
! !

, -200
i

KLV · ·

I

i

I

200

-200

200

-200

,200

-20O
200

recovery phases. The existence of such LT variations in the eastward electrojet location is clearly seen for 1000-1600 UT on April 5. Latitudinal crosssectionsN5 - N8 were specially selected to characterizethe westwardelectrojet dynamics during the transition from the relatively quiet interval prior to the substormonset to the maximum of the substorm expansionphase. During the period of the extremely low Dst valuesfrom 0600 to 0900 UT on April 5 the magnetic field variations along the magnetometer chain are near the quiet level. Such attenuations of the magnetic disturbances along the IMAGE chain are most likely due to the shift of the chain meridian to the day sector where the westward electrojet is locat-

-200
I I I

200
i i i 1

ed poleward normalauroral·.one of latitudes[Akasofu, 19681 . Latitudinal cross sections AX and AZ, throughthe of
eastwardelectrojet at consecutive times are presentUT ed in Figure 2. Arrows mark the location of the center of the emtward electrojet. The equatorward shift of the eastward electrojet center from ..,670 at 1430 UT

-200

OUJ i
200
i
i

!
!

i

I

I

-200

HFIN ·'

Ii
it
I

II
I t
I

20O

!
I

-200
I I

200
I
I

I
I

I
I

I
I I I

I

-200

12

16

20

0

6

6

12

16

20

UT

ú 50 .

05 apr' 1993

Ost

to ..,57Ü at. 1800 UT is distinctly seenduring the storm main phase. During the recovery phase, the electrojet center begins to return to the initial position, but its location at 1430 UT on April 5 is still substantially more equatorward than its location at the onset time of the storm initial phase(1430 UT on April 4). At 1700-1800 UT, AX · 0 in the poleward part of the latitudinal crosssection. This is due to the appearance at theselatitudes in the eveningsectorof the westward directed currents. This corresponds an increasein to the si·.e of the dawn westward ionosphericcurrent cell as described Fukushima by [1953]. However, eastthe ward electrojet doesnot disappearin the eveningsector
as disturbances latitudes increase. Instead it shifts to subauroral The observed shift of with Dst intensification.

0

the eastward electrojet to subaurorallatitudes during storm main phase leads to a substantial underestima-

tion of the AU (AE) indices determined the basis on
-!00

of auroral ·.onelatitude observatoriesonly. The character of latitudinal cross sectionsthrough

Ul

Figure 1. Variations the X component of magnetic fieldalongthe IMAGE chainduringstormmain phase on April 4-5, 1993(at the top). The intensity meawas
sured relative to the quiet level at 0900-1100 UT on

April 4. Dottedlines(vertical) correspond nineunito
versal times with latitudinal cross sections of the AX

the westwardelectrojet, presented Figure 3, and the in eastward electrojet are substantially different. During relatively quiet intervals between substorms,such as 2116 UT on April 4, 1993, and 0132 UT on April 5, 1993 the electrojet center was located at · ... 61Ü near midnight-earlymorning MLT, i.e., in the equatorward part of the auroral ·.one. At the substorm maximums which occur at 2320 UT on April 4 and 0255 UT on April 5, the electrojetcenterdoesnot shift equatorward but the electrojet sharply widenspoleward. The maximum AX field decreaseoccurs at q· .., 64Ü-65Ü, i.e., at
the central latitudes of the auroral ·.one. It is at these latitudes that observatories used for the AE indices cal-

and AJ components presented furtherin the text. The Dst variation the magnetic of fieldfor the same stormis
shown at the bottom. Arrows directed downwards mark

universaltimes of latitudinal crosssections through the culation are placed. Thus, during magneticstormsineastwardelectrojet,and arrowsdirectedupwardmark tervals, the Aoe indicesare correctly determined during cross section throughthe westward electrojet.The loca- the substorm expansivephase maximum and are sigtion of the magnetic midnight eachstationis marked nificantly underestimatedduring the intervals between in substorms. by triangle.


14,22(5
68

FELDSTEIN ET AL- AURORAL ELECTROJETS DYNAMICS

66

62

o ·o
· 56

56

54

qO0
nT

(.
nT

400

8·

nT

04. Oq.93 1430 UT
+· dX, X: dZ

04.0q, 93 1530 UT
+: dX, X' dZ

04.0·,93 1700 UT +' dX, X' dZ

68

66
o

66

· 02
o o

60

(z) 60

c· ·e
(,.)

C. 58
0

56

,
ú

,

,

,

56

_

,
ú

,

ú

ú
ú

. .
ú

54

ú

,

!

i

,

ill

i

!

i

i

mm'!

,

!

,

,

,

54

, i i i ! 1 i i i i i ill

I

I 11111,l

,,m

m,'"l'l

1000

-500
nT

(

5

,.

S00

- 300

0
nT

3()0

6,

04.04.93 1600 UT +' dX, X' dZ

05.04.93 +' dX.X'

1430 UT dZ

Figure 2.

The latitudinalcross sections AX (dottedline) and A2 (solidline) throughthe

eastwardelectrojet on April 4, 1993, at 1430, 1530, 1700, and 1800 UT, April 5 at 1430 UT. Arrows mark the latitudes of the eastwardelectrojet center.

< -150 nT. The auroralelectrojetdynamics duringthe course onesuchstormswerepresented the previous of in section. The EISCAT meridionalchain, the predecessorof IMAGE, recorded moreintensemagneticstorms. All stations of the IMAGE chain began to operate Figure 4 shows variationsof the X component during in Finland in 1993. The chain recorded several modthe main phaseand beginning the recovery of phaseof erate magnetic storms with extremum values of Dst the intensemagneticstorm on May 10-11, 1992. The

Auroral Electrojet Dynamics in an Interval of Intense Magnetic Storm Activity on May 10-11, 1992


FELDSTEINET AL.' AURORALELECTROJETS DYNAMICS
68

14,227

68

66

66

· 82
o) co E

· 62
o)

E
o

·o

58

r.
o

58

56

56

ú

i I iii

i i ii

·1 Iil

!

i-· ! 111

i11

·

100

-50

U nT

50

i

1500-1'000-500

IfilIll IIIII11 IlollS[1 14 illll [1 II1[111 51)0 I ooo
n*i'

04.0t4,93 2118 UT +: dX.X: dZ
68

0·,. 04,93 2320 UT +' dX, X' dZ

66

66

62 o)
E
o

$0

r.
o

58

58

56

ú

56

,,

.,

ú

1!I I ! !I

·[-1tl! I I

i I I t III

5·

.t,00

-200

· nT

2·0

05.0·,. 93 0132 UT +' dX, X' dZ

05,0q,83 0255 UT
+: dX, X: d·

cross sections presented theuniversal are at times quiet of intervals between substorms (April4, 1993 2116UT andApril5 at 0132UT), at therightside-at theuniversal at times substorms of

Figure 3.

Thelatitudinal cross sections through westward the electrojet. theleft side, At

maxima (April4, 1993 2320 andApril5 at 0255 at UT UT). Arrows mark latitudes the the of
westward electrojet center.

ElSCAT

chain did not include the field variations at

Oulujarviand Hankeisalmi, whichare very important whenstudying electrojet the dynamics duringmoderate magnetic storms. intense For magnetic storms with Dst< -200 nT it appearednecessary use observato tionsfrom Russian magnetic stations locatedapproxi-

matelyalongthe ElSCAT meridianup to ·I, ~ eastwardelectrojet shifts equatorwardof the chainduringintense magnetic storms. For the May 10, 1992, storm intense AX > subauroralstations appear immediately after geomagnetic noon and reach a maximum of

51Ü. The EISCAT
0 at the the local 1154 nT


14,228
992
I

FELDSTEIN ET AL.' AURORAL ELECTROJETS DYNAMICS
11 m·¾ 1.992
! I I
i

ed the eastwardelectrojet, are presented Figure 5. in
2O0

The eastwardelectrojet center location is ·I, ~ 540-55 Ü.

-200
I !
!

2OO
-200

The currentdirection changes ½I, 610 (Harangdisat ~ continuity). The appearance this discontinuity of at suchlow latitudesand early MLT (~13) occurs only duringintervalsof intensemagneticstorms.
Figure 6 showslatitudinal crosssectionsAX and AZ
for times when the westward current was located above

200
-200
KIL
t i
i

i

200

MUO
t i i i 1

200 · -200 n c·

i

i

i

200 ·
x

all the chain stations. The absence Oulujarvi and of Hankasalmi data prohibitsthe reasonably precise localization of the westwardelectrojet center. However,the field variations clearly show that at the times of maxima substormintensity the electrojet centeris located at higher latitudes than during the intervalsbetween substorms in the beginningof the expansion or phase
of substorms at auroral latitudes.

-200
NUR
200

Discussion

-200
i !

LNN
t i

,
i

200 -200

During the storm main phase it is necessary take to
into account data from subaurorM observatories for the

BOX
i i i i

AE indicescMculation.The usageof corrected AE in-

200
I I

i

i
i

i
i

-200

dices previously has shown [Feldstein al., 1994], et that (1) duringthe stormmain phaseenergy fluxesin the
ring current and those of the aurorM ionosphere in-

MOS

·, ·
i

I i i I

I I

I

I

I

I

I

I
1

2O0 -200

I

I

6

1o

1·,

18

22

2

UT
-50

crease simultaneously (2) duringthe stormmain and phasethereexistsclose relationship between geoefthe feetire solarwind parameters and energyflux entering
in the magnetosphere. The eastward electrojetdynamics duringgeomagnetic stormswas investigatedby Kamide and Fukushi·na

Der
n!

-100

[1972], Grafe[1983], and other. As a rule,the electrojet
-150

J

-200

centershiftsequatorwardduring a substorm but practicMly alwayslies within the auroral zone at ·I, ~ 630-65 Ü. The centershift during the magneticstorm of April 4, 1993, is shownin Figure 2. For the determinationof the eastwardelectrojet center within 10 min intervals, a modelof a homogeneous current layer at 120km alti-

;

:

:

6

; I0

:

:

:

; I·,

;

:

:

:: 18

:

:

:: 22

:

;

; 2

:

:

:

-300 UT

tudeabove Earth'ssurface the [Grafeet al., 1987]was

used. The resultsfor every 10-min interval are depicted in Figure 7 by a solid line. The electrojet center Figure 4o Variation of the magnetic field compograduallyshiftsequatorwardup to ½I' 570 at 1800 UT ~ nent along the EISCAT chain dufin· the m·n phase duringthe storm main phaseand then, when Dst stabiof the stormon May 10-11,1992(at the top). The intensity was me·u·ed ·elative to the level at 0600-0800 lizes at the level of -100 nT for a short while, returns ~ phase, UT on May 10· 1992. Variations of the X components to ½I, 600 . At the beginningof the recovery when Dst ~ -100 nT, the electrojet centeris locatedat a·e added fo· Leningrad, Bo·ok, and Moscow.The D· wfiation of the magnetic field fo· the same storm is ½I'~ 60 Ü. The latitude of 600 should thus be the locap·esented the bottom). Notationsa·e simfla· to tion of the'eastwardelectrojet centerfor the stable Dst (at ·i·u·e 1. level with intensity ~ -100 nT. The joint influences of the ring current and substormsshiftsthe eastwardelectrojet center somewhatequatorward, which is observed
]0 may 1992 I I may [992

at 1030 UT. In the auroral zone, where AX is usually positivein the eveninghours, the westwardcurrents were recordedduring this time. Such currentsflow in the auroral zone during the whole storm main phase. Hence,if only the region of auroral latitudes is considered, the impression can be created that the eastward electrojetdisappearduring the courseof intensivemagnetic storms. Latitudinal cross sections AX and
when subauroral and midlatitude observatories record-

at 1800 UT.

A more sophisticated model of the ionospheric current distributionbasedon the magneticfield measurementsalongthe meridian chain wasusedby Popovand

Feldstein [1996]. The latitudeinterval above magnethe
tometer chain was divided into 50 infinitely thin bands layer at 115 km altitude. Current bands are assumed to be infinitely long, and the currentdensity in every j band is homogeneous determinedby Biot-Savart's and


FELDSTEIN ET AL.' AURORAL ELECTROJETS
68

DYNAMICS
, , ,

14,229

,

,

,,

\

:

\.
\
·i,,ii',,', ',,,'·,',,,·, ,,,,,,',,,
.

i
ú

ú

..

'

i i i i i ·1i i i

I i i t i i i i"1

i i i i'1 I'1 i i

BOO

·

600

1, 7oo

·00

l:i nT

5·)0

!

nT

nT

+," dX, X:

dZ

10.05.92 1218 UT +: dX, X: dZ

10,05.92 1514 UT +: dX, X: dZ

Figure 5. Thelatitudinal cross section (dotted X line)andg (solid line)through eastward the
electrojet May 10, 1992,at 1034,1218,and 1514UT. Notations similarto Figure2. on ate

To obtain a detailed structure of the meridian 1992].The numerals near the circles UT hoursfor are cross sectionof j, i.e., currentvaluesat 50 points,the the March 23-24, 1969, strong magnetic storm. The inverse problem solved the regulari·.ation is by method straight line has been obtained by the least squared [Tikhonou Arsenin, and 1974]. The modelallows the method. The electrojet movesto the lower latitudes as and its positionis describedby the relaseparation the observed of groundmagnetic field vari- DR increases law.

ations into internal and external sources. Figure 8

tionship

presents latitudinaldistribution external(ionothe of ú -- 65.20 -'10.035DR (1) spheric) source currents calculated usingthe method developed Popou Feldstein by and [1996].The eastward in the 0 > DR > -250 nT interval, where DR is in nancurrents spreadovera large rangeof latitude. Arare
rows mark 'the latitudes of the current maximum, which

oteslas.Crosses Figure 9 correspond the westward in to electrojet center in the intervals betweensubstorms for the April 4-5, 1993 and May 10-11, 1992, magnetic shiftsequatorward the magnitudeof Dst increases. as The latitudinal profilesallow the integral current val- storms. Good quantitative agreementof data for all storms clearlyseen.Thus the data of is ues to be calculated. For example, integral valuesof threemagnetic the eastwardcurrentin the Figure 8 cross sections, are the IMAGE and the ElSCAT meridian chainsof magsupportthe earlierobtainedrelationship be61.5x 103Aat 1430 UT, 417.gx 103Aat 1530 UT, netometers 476 x 103 A at 1700 UT, 597 x 103 A at 1800 UT tween the westward electrojet center position and the

on April 4 and 525 x 103 A at 1430 UT on April DR or Dst intensity.
During the storm main phasethe centerof the west5. The comparison the currentintegral valuesand of the currentmaximumvaluesalongthe latitudinal cross ward auroral electrojet shifts equatorwardto · ~ 580 sections showsthat they changeasynchronously. The when Dst ~ -200 nT and ·I, ~ 540 when Dst ~ -300 characterof the magnetic integralcurrentvalues take into account only the nT. Assumingquasi-dipole not it intensity, but the electrojetwidth, and, therefore is field lines i'n the inner magnetosphere, meansthe loit moreprecise characteristic disturbance of level,accord- cation of the electrojet at oe shellsof ~3.6 and ~2.9,

ing to KamideandAkasofu [1974], than A U or AL indices.

respectively. Lyons[1996]suggest that the processes,
connectedwith the electrojet formation deeply in the

can Figure 9 shows the locationof the westward elec- inner magnetosphere, be the causeof energeticion trojet centerduring the main phaseof the magnetic fluxeswhich form the low energypart of the ion spec-

storm in the near-midnightto early dawn MLT sector tra in the inner part of the ring current. These ions for intervalsbetweensubstorms versusDst [Feldstein, are mainly oxygenfrom the upper atmospherewhich


14,230
66

FELDSTEIN ET AL.' AURORAL ELECTROJETS DYNAMICS

66

'o

:3 64

....,

o)
(0

E'

O 56
(3) '

0 5z,
·2

50

,!
n

0

t50

:)

nT

n"

I0.05.92 t9i6 +: dX, X: dZ

UT

10.05.92 2059 UT +: dX, X: dZ

11.05, 92 0009 UT +: dX, X: dZ

66
'

i

i[1 i

i,

i

i

i i,

ii

i

i

i

ii

i

ii

i iiii

i!

.-

X
_

i
Ii

i

i

:·+

,,,/
\

_i
ú .
I .

,

IL

:
:
c
E.

,/
.

6O

i

' I/
ú
.. .

O: 50

.
ú

ú
ú

\

:

· c: 56:
o 54
52
..

\+
i

-

i

i

.
ú
ú

."
ú

:
,-

..

ú

:

.

i I ·11 i i iiII

i i i ii1 i i i i

IiiiIiIII

iiiiiiiii

111111111

'llll,1111

1111111111

,1 ooo

-500

0

5

nT

1ooo -500 !l
n

5)0 1

1500- t000-500

nT

0

5(.)0

1,

oo

10, 05, 92 i 6:]9 UT
+., dX, X', dZ

10.05.92 2t26 UT +: dX, x; d!

11,o5,92 0105 UT +,' dX, X: d[

Figure 6. The latitudinalcross sections throughthe westward electrojet duringintervals between substorms in the beginning a substorm or of expansion phasein aurorallatitudeson May 10, 1992,at 1918, 2059 UT and on May 11 at 0009 UT and at substorms maxima on May 10 at 1839,2128 UT and May 11 at 0105 UT. Notationsare similar to Figure 3.


FELDSTEIN ET AL.' AURORAL ELECTROJETS DYNAMICS

14,231

70·._.

magnetic field (IMF). In this situationa change the in
IMF B, magnitude doesnot influencethe AE intensity, i.e., we have the saturation effect in the AE indices. The maximum of the AE indicesis related to the sharp poleward widensof the westward electrojet. The chain of the AE observatories begin to record the electrojet with the ma0dmum intensity in close connection with the magnitude of the IMF southward component,i.e.,
the saturation effect of the AE indices is absent.

Kamide [1979] discovered interesting an peculiarity in
centreof eastward electrojet

t ,t,I

t_ I
20,00

the behaviorof the A U and AL indicesduring magnetic storms, namely, A U - AL during the initial phase and AU << AL during the main phase. This peculiarity

of A U (AL) indices a consequence the electrojets is of

dynamicsin the courseof a magnetic storm. At the initial phase, when the ring current is week, the elecApdl 4, 1993 (UT) trojet centersare located at AE observatories latitudes and, therefore, A U = AL . During the main phase Figure 7. The eastwardelectrojet center shifts for the electrojetscentersshift equatorward of the AE obthe stormmain phase April 4, 1993(solidline). The servatories on latitudes, though the westwardelectrojet sicrossmarks the center position at 1430 UT on April multaneouslywidens poleward to the AE observatories 5. Arrows mark univars·l times of the latitudinal cross latitudes. Suchdynamicsof the electrojetsexplainsthe sections Figure 2. in

peculiarity A U (AL) indices of behavior described by Kamide[1979].

can be acceleratedup to ~10 keV energydirectly from the ionospheric heights. However,the bulk ion population in the ring current has energy in the 50-100 keV interval. Apparently, ions of such energiesare accelerated from the ionosphere higher latitudes during inat tense substorms,when the westward electrojet widens poleward and its maximum intensity at the substorm

The above presentedanalysis of the magnetic field variationsduring magnetic stormsintervals showsthat the eastward electrojet does not disappear when an

intensive magneticdisturbance develops (this conclusions yard for other stormsas well). For the magis

netic storm on May 10, 1992, the eastward electrojet, when Dst ~ -300 nT at 1500-1600 UT, was observed at subauroral and midlatitudinal stations Nurmijarvi, maximum occur at the central latitudes of the auroral Leningrad, Borok, and Moscow, where the variations zone. From thoselatitudes ionsof ionospheric origin en- AX > 0 reached ~1100 nT for Borok. The positive existed for a long ter the plasmasheetand then moveearthward(large- valuesof AX at these observatories scalemagnetospheric convection).It is possible that interval during the storm main phase,when intenseDst the ionosphericion flux toward the plasma sheet at the was accompaniedby very intense eastward electrojet. substormma0dmum substantiallylarger than that to- It is quite possible,that conclusionon the existence of is ward the inner magnetosphere during intervalsbetween caseswhere there is no eastward electrojet anywhere substorms. When convectingearthward, allowing for during the main phase when the westwardelectrojet is the first adiabatic invariant conservation,ions acquire very intense[Karnideet al., 1976; Karnide,1979]readditional energy and constitute the main population quires additional more careful substantiation. At any of the ring current with energies severaldozenskeV. rate data from the meridian chains IMAGE and EISof The electrojetlocation during the storm main phase CAT showsthat the eastward electrojet exists in the and their dynamics in connection with substorms al- eveningsectorin every magnetic storm. The electrojet low to interpret someeffectsdescribedin the literature. shifts equatorwardof auroral zone latitudes occupying Among them is the AE saturation effect describedby more and more early hours with increaseof the Dst inWeimeret al. [1990]. It occurred whenB· is south- tensity. The equatorward shift of the electrojet duringmagward with a large magnitude, i.e., during storm main phase. It was assumedthat the AE saturation is the netic stormsshouldexist not only along the meridichains. result of the nonlinear nature of the magnetosphere- an of EISCAT-IMAGE magneticobservatories

the of et ionosphere coupling [Kan et al., 1988]. On the basis Therefore, inference Weirner al. [1990,p.18,984] of electrojet dynamics discussed above for the storm that"byexamining magnetic records College from (·I, Ü) Ü) main phase,the followinginterpretation can be offered 65 and Sitka (· = 60 duringthe International for the AE saturation effect for AE minimum values Geophysical Year (one of the most disturbed periods and its absence for AE maximum values. For small over recorded the century), we found that the 2 in from Sitka almost always showsnegative AE values, i.e., during intervals between substorms, component' indicatingthat the westward electrojetwasloboth the eastward and westwardelectrojets are located changes,
equatorward of the chain of observatories intervals cated poleward of Sitka," requireson additional confor of large southward B· componentof the interplanetary siderationusingdata from Sitka.


14,232
68

FELDSTEIN

ET AL- AURORAL ELECTROJETS

DYNAMICS

66

'
* , ú

=

ú

ú *
ú

ú

ú
ú

o

ú

·

,6o

i. I
ú ú

I Ii i
$
ú

!_

ú

(o

58

.
_
.

,,

ú

ú ú
.

ú

.

ú

,

54

i!111111

FI/m 04.OA, I q30 UT 93

-1

I

2

-2

R/m 04.04.93' 1530UT

-1

ú

i1(11·111

iiii11111

iiiiiiiii.

0

!

2
R/m

04.04,03

1700 UT

68

_

m

i

iml

m

ii

ú

68

ú

,,

ú ú

'6B'G·

ú

'

'

66

'

ú

$

E'
O'

·

60:

'
ú

ú

,,-

,

,

ú ú

E
O

·
ú

60

ú

(- 58''
0 ' *
,. ,,

' ©

'
ú

'

0

(- 58

56 ' .
,,

ú

'

56

ú
ú
ú

ú

5·
_9

...... ', ,, ,', .... "',,, ·,..............
-I (t
m

I

R/m

04,04.93

1800 UT

05.04.93

lZ·30 UT

Figure 8. Thelatitudinal cross section thecurrent of density through eastward j the electrojet fromtheexternal sources [Popov Feldsiein, and 1996] themagnetic for storm April 4-5, 1993. on
Arrows mark the current maximumposition.


FELDSTEIN

ET AL'

AURORAL ELECTROJETS

DYNAMICS

14,233

2f.
-'

·",·

25

x00.09

Proper disturbance for such a consideration is the storm with main phase on September 29, 1978. The maximum values of Dst ~250 nT fall on ~1200 UT, i.e., near midnight hours for Sitka, when the intense westward electrojet is expected at the station meridian. Figure 10 presentsfor the period between September 28 and September 30 of 1978, the interplanetary medium parameters variations, the AE and Dst in-

'

01.32 X ·o 20.59

dices, and hourlymeanvalues horizontal of (AH) and vertical(AZ) components the magnetic of fieldfor Sit-

Figure 9. centersin the near-midnight-earlydawn MLT sectoras the function of DR or Dst intensity. The numerals at the circles and crosses are UT. The straight line has been obtained by the least-squaresmethod.

ward currentwith centerpolewardof Sitka (AH < 0 and AZ < 0). On September duringthe stormmain 29 - 100" -200 Dst, DR, nT phase,the westwardelectrojetcenter(AH < 0) was of The positionof the westwardelectrojets locatedinitially poleward Sitka (A· < 0) and then,
as Dst variation intensifies, the electrojet center shifted

24 19'18 ··

ka. On September 28, 1978, the substorm with the AE intensity up to 1000 nT is characterized by the west-

to the zenith (A· ~ 0) and then farther equatorward (A· > 0) of station.Thusthe data presented Figure in
10 indicate that intense westward electrojet is located

Ill/Ó
,

30

Bz
nT

0,
-313

Figure 10. Variationsof the solarwind speed,the Bz and IBI of the IMF, indicesAE and Dst, and Sitka'svariations the AH and A· components deviations of (as from quiet day field values} the stormon September for 28-30,1978.


14,234

FELDSTEIN ET AL.' AURORAL ELECTROJETS DYNAMICS

! ooo

poleward Sitka (·I, = 60Ü) near midnightbeforethe of
main phase of the intense magnetic storm but shifts equatorwardof the station during the main phase, The storm on September 29, 1978, is not unique, A similar situation was observedfor the storm on August 28, 1978, as well. For this storm data on the interplanetary medium parameters,the geomagnetic activity indicesand hourly mean valuesof the magneticfield variations for Sitka are shown in Figure 11. The westward electrojet center shifts to the station zenith as Ost developsand then moves equatorward of the sta-

v
km/e

500

Bz

0

-30

tion (Aft < 0, AZ > 0). Thus,duringthe mainphase magnetic of storms (Dst · -250 nT) the onehour averaged westward electrojet
is located equatorward of Sitka. Hence, near the Alaska meridian the westwardelectrojet center averaged for

30

/B/is
nT

I hour is located equatorwardof ·I, = 600 during the main phase of intense magnetic storms as well as over
EISCAT-IMAOE,

o

Acknowledgments. We are grateful to ElSCAT magnetometer and IMAGE teams for possibility to use magnetic observatories chain data. This research was supported by the Russian Foundation of Basic Research, grants 96-0566279 and 96-05-65067, INTAS-RFBR-95-0932.
The Editor thanks S.-I. Akasofu and R. Nakamura for

7500
5O

their assistance evaluating this paper. in
,,

References
Akasofu, S.-I., Polar and magnetosphericsubstorms,D. Reidel, Norwell, Mass., 1968. Akasofu, S.-I., Relationshipsbetween the AE and Dst in-

dicesduring geomagnetic storms, jr. Geophgs.Res., 86,

-250

2OO

4820-4822, 1981. Akasofu, S.-I., B. H. Ahn, Y. Kamide, and J. H. Allen, A note on the accuracy of the auroral electrojet indices, jr. Geophgs. Res., 88, 5769-5772, 1983. Allen, J. H., Auroral electrojet magnetic activity indices

(AE) for 1970,Rep. UAG-,·,·,World Data Cent. A Solar
Terr. Phys., Natl. Oceanic and Atmos. Admin., Boulder, Colo., 1970. Allen, J. H., C. C. Abston, and L. D. Morris, Auroral elec-

·z
nT
-200

trojet magneticactivity indicesAE (11) for 1973, Rep.
UAG-47,.World Data Cent. A Solar Terr. Phys., Natl. Oceanic and Atmos. Admin., Boulder, Colo., 1976. Chapman, S.C., and J. Bartels, Geomagnetism,Clarendon, Oxford, 1940. Davis, T. N., and M. Sugiura, Auroral electrojet activity index AE and its universal time variation, jr. Geophgs. Res., 71, 785-803, 1966. Feldstein, Y. I., Modelling of the magnetic field of magnetospheric ring current as a function of interplanetary medium, SpaceSci. Rev., 59, 83-165, 1992. Feldstein,Y. I., A.E. Levitin, S. A. Golyshev, L. A. Dremuhina, U. B. Veshchezerova,T. E. Valchuk, and A. Grafe, Ring current and auroral electrojetsin connectionwith interplanetary medium parameters during magnetic storm, Ann. Geophgs.,1,·, 602-611, 1994. Fukushima, N., Polar magnetic storms and geomagnetic bays, jr. Fac. Sci. Tokyo Univ., 8, 293-412, 1953. Gonzalez, W. D., J. A. Joselyn, Y. Kamide, H. W. Kroehl, G. Rostoker,and B. T. Tsurutani, What is a geomagnetic storm? jr. Geophgs.Res., 99, 5771-5792, 1994. Grafe, A., Electrojet boundariesand electron injection boundaries, jr. Geomagn. Geoelectr.,35, 1-15, 1983.

-·oo

-600

-BOO

- 1ooo

o

12 [iT

o

Figure 11. Data similar to presented Figure 10, in but for the stormon August28, 1978.


FELDSTEIN

ET AL.: AURORAL ELECTRO JETS DYNAMICS

14,235

Grafe, A., R. J. Pellinen,W. Baumjohann, and M. Vallinkos- Kan, J. R., L. Zhu, and S.-I. Akasofu, A theory of substorm: onset and subsidence,J. Geophys. Res., 93, 5624-5640, ki, Development the auroral electrojetson 16 March of
1978: An event study, Geophysica,83, 113-141, 1987.
1988.

Kamei, T., and H. Maeda, Auroral electrojetindices(AE Lyons, L. R., Round-table discussion,paper presented at Chapman Conferenceon Magnetic Storm, AGU, Pasade) for January-June, 1978,Data Book3, Fac. Sci., Kyoto
Univ., Kyoto, Japan, 1981.

Kamide,Y., Relationbetween substorms storms, Dyand in namicsof the magnetosphere, edited by S.-I. Akasofu,pp.
425-443, D. Reidel,Norwell, Mass., 1979.

na, Calif., Feb. 12-16, 1996. Popov, V. A., and Y. I. Feldstein, About a new interpretation of "Harang discontinuity", Geomagn. Aeron., 36,

Sumaruk, P. V., Y. I. Feldstein, and B. A. Belov, The dynamics of magnetosphericactivity during an intense magthe aurorM electrojet and its relation to the interplanetary netic storm, Geomagn. Aeron., $9, 110-115, 1989. magnetic field polarity, J. Geophys. Res., 79, 3755-3771, Tikhonov, A. N., and V. Y. Arsenin, Methods of noncorrect 1974. problems solution, Moscow, Nauka, 1974. Kamide, Y., and S.-I. Akasofu, Notes on the auroral electro- Weimer, D. R., L. A. Reinleitner, J. R. Kan, L. Zhu, and jet indices, Rev. Geophys.,$1,1647-1656, 1983. S.-I. Akasofu, Saturation of the auroral electrojet current Kamide, Y., and N. Fukushima, Positivegeomagnetic bays and the polar cap potential, J. Geophys. Res., 95, 18981-

Kamide, Y., and S.-I. Akasofu,Latitudinal crosssectionof

in eveninghigh latitudes and their possibleconnection with partial ring current, Rep. Ionos. SpaceRes. Jpn,
$6, 79-101, 1972.

18987, 1990.

Kamide, Y., H. W. Kroehl, M. Kanamitsu,J. H. Allen, and 142092 Troitsk, Moscow Region, Russia S.-I. Akazofu, Equivalent ionosphericcurrent representaA. Grafe, GeoResearch Center, D-14482 Potsdam, Gertions by .a new method, illustrated for 8 - 9 November many 1969 magneticdisturbances,Rep. U'AG-55,pp. 1 - 11, World Data Cent. A Solar Terr. Phys., Natl. Oceanic (Received April 26, 1996; revisedOctober21, 1996; and Atmos. Admin., Boulder, Colo., 1976. accepted February17, 1997.)

Y. I. Feldstein, L.I.Gromova, and V. A. Popov, IZMIRAN,