Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.izmiran.rssi.ru/magnetism/Feldstein_2012/PDF/262.pdf Äàòà èçìåíåíèÿ: Wed Mar 11 16:43:46 2009 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 21:06:20 2012 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: ï ï ï ï ï ï ï ï ï ï

GEOPHYSICAL

RESEARCH LETTERS,

VOL.

10,

NO.

8,

PAGES 749-752,

AUGUST

1983

A STATISTICAL
OF THE

STUDY OF THE DYNAMICS OF THE EQUATORWARD BOUNDARY
DIFFUSE AURORA IN THE PRE-MIDNIGHT SECTOR

J A. Sauvaud J Crasnier 1 Yu I 1,

Galperin2, Y I

Feldstein 3

1C.E.S.R., CNRS-Universit· Paul Sabatier, 31029 Toulouse, France 2SpaceResearch Institute (IKI), USSR Academy Sciences, Moscow,USSR of 3IZMIRAN,USSR Academy Sciences, Moscow of Region, USSR
Abstract. We present a statistical study of the dynamics of the equatorward boundary of the diffuse aurora. This Soft Electron Boundary appears to be best fitted by using the AE index values averaged over the 5 hours preceding the measurements. This inertia of the boundary is interpreted as indicative (1) of the action of a polarization electric field braking the inward motion of the the plasma sheet would correspond, for stationary conditions, to the boundary of the "forbidden" region for inward-moving plasma (Alfv·n Layer) which is determined by the strength and pattern of the convection electric field. Such an assumption has implicitly been made in order to derive the equatorial electric field strength from particle measurements of the plasma sheet inner boundary

correspondingequatorial plasma during increasing
magnetic activity so that several successive perturbations are needed to obtain a consequent inward
displacement and (2) of a weak wave-particle interaction during periods of decreasing magnetic activity.
Introduction

[see e.g. Hultquist et al.,19821or of the equatorward boundary of the diffuse auroral zone [Kamide and Winningham, 1977]. For non-stationary conditions, i.e. if the characteristic time for electric field variations is shorter than the particle lifetime, the instantaneous boundary position will
depend on the past history of the geomagnetic activity (so that no information on the convection electric field can reasonably be deduced from the instantaneous boundary location).On the other hand, in the case where particle lifetime is shorter than

After the detection of the existence of a permanent·i. broad, and rather uniform diffuse auro-

ral emission [Lui and Anger, 1973], simultaneous
particle observations at geostationary orbit and near conjugate field lines have shown that the spectral shape and differential fluxes of precipitating electrons responsible for the diffuse aurora are very similar to those of the trapped plasma

the characteristic

convection time, the boundary

position, even for stationary conditions, is no longer fixed by the pattern of the convection electric field: the corresponding flux tube is empty of plasma sheet particles when it reaches the convection boundary of the forbidden region for

sheet electrons [Menget al., 1979].This result
has led to the conclusion that the diffuse aurora

inward-moving plasma. As the plasma is rapidly lost
from the inward-moving flux tube we can expect a

is produced by a direct
electrons. It

dumping of the plasma sheet
that the electron

weak inertia
conditions.

of the boundary under non-stationary

has been suggested

cyclotron harmonic waves detected near the equatorial plane aboard OGO-5 and IMP-6 are at the origin of the strong pitch angle diffusion of ·keV plasma sheet electrons into the atmospheric loss cone and of their subsequent precipitation into the auroral

This paper presents a statistical study of the dynamics of the equatorward boundary of the diffuse auroral zone in the 22-24 MLT sector as a function of the past history of the geomagnetic activity evaluated from the AE values averaged over

ionosphere [Lyons, 1974]. However, this generally accepted view has been challenged recently by measurements of electrostatic electron cyclotron waves performed aboard the GEOSsatellites showing that more than 85 %of the time the minimum wave elactric field amplitude for strong diffusion is not reached

different time periods preceding the boundary crossings. This S_oftElectron B_oundary (SEB), considered to be the inner edge of the polar flux tubes best organized by ionospheric projection of the plasma clouds injected into diduring substorms, appears to be using AE values averaged over the

[Belmontet al., 1983]. Belmontet al. [1983] conclude from their analysis that these waves are not the only cause of diffuse electron precipitation or alternatively that the wave particle interaction

5 hours preceding the measurements, indicative of
a strong inertia under non-stationary conditions.
This result implies that the convection characteristic time is short compared to the particle lifetime and shows that the inner magnetosphere dynamics cannot be studied without taking into account the past history of the geomagnetic activity, i.e. of the past solar wind- magnetosphere coupling. Tentative correlations between the SEB location and the intensity of the IMFBz component,

It

is weak. must be

stressed

that

the

nature

of

the

wave-particle
vection electric

interaction,
field,

together with the concontrol the nature and

dynamics of the earthward termination of the plasma sheet inner edge or equivalently of its ionospheric projection: the equatorward boundary of the

although carried out on fewer samples, indicate

diffuse auroral zone[seeSouthwood Wolf,1978]. and
Two extreme casescan easily case of a particle lifetime
characteristic convection

the same result.
Data Selection

be conceived : in the much larger than the
time the inner edge of

Copyright 1983 by the American Geophysical Union.
Paper number 3L1032. 0092 -& 276/83 / 003 L-1032503.00
749

The equatorial SEB was determined using in situ measurements of auroral electrons in the energy

range 0. 2 -16 keV made aboard the non stabilized
AUREOL-1 and AUREOL-2 polar satellites. On each spacecraft three identical spectrometers were used:


750

Sauvaudet al.'

EquatorwardBoundaryof the Diffuse Aurora

1091·1500 JANUARY197·. I 1 22, 2 ! t I?·./01/22 AUREOLE·1000 AE
· 108 6 12U.2/,_ -5000 T 18
· 1071
2·

tical

From the AUREOL-1 and 2 satellites the analymodel of the SEB invariant latitude as a

function

of the magnetic

local

time and the Kp
2

·, 500

Kp / = 2o
1

index is given as follows[Galperinet al., 1977] :

AOSEB = 71.47- 1.25 Kp- 0.018Kp

-(2.84+1.24Kp0.076Kp (--g---3) (2) 2)
Figure 2 illustrates a comparison of these two models. The agreement is exceptionally good, with
a latitudinal difference between 0.1 and 0.7 Ü for

MLT

....106

boundary the 6300%diffuse aurora,determined of
from ground based photometer data, and the SEB
location also
that

a Kp index higher than 1. Itmust be stressed that comparison between the location of the equatorward
shows an extremely good agreement
the AUREOL

·, 10 5

0.·2ú _
1.70 key

[Slater et al., 1980]. From these correlations we
conclude the SEB measured onboard

lO ·'
·
19:56 ?5.2 21.5 :5? 72.? 22.1 :58 69.8 22.5 :59 66.9 22.8

satellites
the diffuse
-

represents the equatorward boundary of
auroral zone with a high degree of

3.60keV

confidence.

_ 16.kj·V :·0
20:00 63.9 23.0 :01 U.! 60.8 23.1

Dynamics of the Boundary

The fact that the equatorial boundary of the diffuse auroral zone can be organized by the 3 hour

Kp index is already a strong indication

that this

·, ], 9·e·e·C·a· e·ecC·o· ·ux ·o· selected e·e·es a·o· a saCe·Ce pass ·om the po·e·a·d bou·da·7 o· the d·sc·ece au·o·a· ova· ·o the equaco·a·d bou·da· o· the d·se au·o·a· ·o·e,

boundary is stable over periods of several hours. However, this magnetic index, computed for fixed UT intervals, does not allow a detailed study of

twoviewing opposite directions,the third viewing vity which known controlthe plasma is to sheet at a 90 angleto them. this study have Ü For we taken dynamics [Arnoldy Chan, and 1969]. this study In the
into account the SEB crossing in the 22- 24 MLT sector and in the altitude range 400-1000km for

the boundary dynamics. The 2.5-minute AE index appears adequate to perform this kind of study as it can be averaged over any chosen time period and furnishes a measure of the ongoing substorm acti-

pitch angles inside or close to the loss cone. The SEBposition is defined by a threefold increase of
the 1- 2 keV electron flux which, except during

AE index has been averaged over the 1/3,2,3, 5, 10 and 20 hour periods preceding the exact time of the SEB crossing in the 22- 24 MLT sector. Figure 3 presents the invariant latitude of the SEB as a
function of the various averaged values of the AE

substorms,is usually displaced by 1Ü- 2Ü poleward
of the softest (0.2keV) electron precipitation.

Figure 1 illustrates
flux variations

some differential electron
70

along a typical AUREOL-2 crossing

of the auroral zone from North to South in the 22 MLT sector. These measurements refer to the end of

the expansionphase of a substorm with an amplitude
of about 300¾, around20:00UTon January 22, 1974.
The northern boundary of the auroral
to discrete auroral

electron

pre-

c·
z

66

cipitation is well defined and probably corresponds
forms which are known to reach

their highest latitudes

at this epoch of a substorm
o 62

[StarkovandFeldstein, 1971 The equatorward ]. boundaryof the soft electron precipitation exhibits an energy-dependent structure with the lowest

energy electrons precipitating at the lowest latitudes; however, the 0.22 keV and 1.70keV electrons
show the same latitudinal profiles at the SEB. The
may obviously
In order to
c·

58

SEBlocation is indicated by an arrow at Ao=65.3Ü
and MLT = 22.9 H. The SEB location
teria selected for its
56

depend the detector sensitivity andon the crion
determination.

test the validity

of our SEBdetermination we have

compared location of this boundary the independently using the AUREOL-2 DMSP/F2 and data sets. Fromthe DMSP/F2 satellite the fit to the SEB as a function of Kp in the 22-23 MLTsector gives [Gussenhoven al., 1981] : et

Fig. 2. Latitude of the equatorward boundary of the diffuse electron precipitation as a function of the Kp index. Comparison betweenthe AUREOL-1/2 results and the DMSP/F2 results. Statistical location of main ionospherictroughdetermined by
Kohnlein and Raitt [1977] is indicated by a dot-

AOSEB - 1.79 = 68.3 Kp

(1)

dashed line.


Sauvaud et al.'

Equatorward Boundary of the Difoeuse Aurora

751

7O

I
--

I I I 11111

I

I I I Illl

I

I I I I III I
AE

I

I III
2 HOURS

I

I I I I I!1 I
AE

I
,

I I I I I!!
3 HOURS

--

AE, 20MINUl'ES
68
66 ·64 _
coo ·
·e
ú

o
-w

60
58

* .·

ú

o
'-

S6
S4

r2= 0.77

SE 1.67 =
AE, 5 HOURS

·

r2=0.$1 SE 1.53 =
AE
,

·
·-

70
68

10 HOURS

AE, 20 HOURS

<
·
z

66
64
62

ú

-

x,

ú

ú
6O

.\

58

\.

ú .·r
2

ú ·e

r 2:0.90 ·
lO

55 =1.13
I I I I I iii

= 0.76

SE:

1.7·.
ú

r
I i I Ill

2

= 0.61

SE:

2.05
ú -

· ,,,,,,I
lOO

,
lOOO

, ,,,,,,1
lOO

I

I

·

,

, ,,,,,,I
lOO

,

, ,,,,,,
lO!)0

lOOO

AE

INDEX

(GAMMAS)

Fig. 3. Dependence of the SEB on the AE index averaged over 20 minutes, 2, 3, 5, 10 and 20 hours before each crossing. Dashed line indicates a quadratic fit determined by the standard least squares method.

index. The general trend is a boundarymotion toward a lower latitude when the magnetic activity

quently displace the earthward termination of the plasma sheet inner edge closer to the Earth rela-

is prolonged and gradually increases. In each case the dependence nonlinear and can be approximais
ted as :

tive to its quiet time position and, on the other hand the precipitation boundary, though less intense, can still be observed for several hours
position reached even when the

Inv. lat.(deg) A+Bln(A-·)+C[ln = (·)]2
This fit
method and is shown by a dashed line. Table ]

(3)

magnetic activityis decreasing.several In cases
this weak and remanent precipitation which still produces an observable enhancement of the F-layer ionization can be simultaneously observed with a new and more intense precipitation, due to more recent injections, located at higher latitudes
TABLE ]. Result of the fitting procedure various average AE values. for

near

the lowest

was determined by the least squares
the square of the correlation
conclusion which can be derived from

summarizes the values of the three parameters (A,
B, C) of Eq. 3,
The first

coefficient (r 2) and the standard error (SE).
these numbers is that the boundary position is best organized by taking the AE index averaged over the 5 hours preceding the boundary crossing. The corresponding correlation coefficient is furthermore

Averaging period A
20 minutes
2 hours
3 hours 5 hours

B
, ,

C
-0.54
-0.78
-0.88 -0.75

r2
0.77
0.81
0.86 0.90

SE
._

higher than in the cases of fits
the hourly

[Galperinet al.,]977; Gussenhoven al., ]98] ] or et
value of the Bz component of the inter-

using the Kp index

64.81
60.69
59.29 64.1]

2.61
4.61
5.40 3.7]

1.67
1.53
].33 1.13

planetary magnetic field [Kamide and Winningham,

]977]. This result indicates ·hat the boundary
location dependsmainly on the/ past magnetic act' vity and apparently presents · strong inertia which can schematically be described as the consequence of two effects: a strong magnetic activity sustained over several hours is necessary to conse-

]0 hours
20 hours

57.53
81. ] ]

6.4]
-3.54
,

-1.02
0

0.75
0.6]

].74
2.09


752

Sauvaud et al.'

Equatorward Boundary of the Diffuse Aurora

[Khalipov et al., 1977; Valchuk et al., 1979]. During
disturbed periods the five hour period includes several substorm-associated plasma injections toward the inner magnetosphere. Note that this statistical study does not take into account probable radial motions of the boundary in the course of individual magnetospheric perturbations.
Discussion and Conclusion

in the magnetosphere, J. Geophys. Res., 76,
_

3587,

1971.

Galperin, Yu.I., et al., Auroral diffuse zone. Part I: Model of the equatorward boundary of the diffuse zone of auroral electron precipitation in the premidnight and midnight sectors, Cosmic
Res., 15, 362, 1977.

Gussenhoven, M. S., et al.,
tions their

DMSP/F2 electron

observa-

of equatorward auroral boundaries and relationship to magnetospheric electric

During weak magnetic activity periods, ponding to near stationary conditions, the memory of the equatorward boundary of the aurora implies a weak pitch angle diffusion low energy electrons forming the boundary:

corres5-hour diffuse of the at L= 6

fields, J. Geophys. Res., 86, 768, !98!.
Horwitz, J.L., et al.,On the relationship of the plasmapause to the equatorward boundary of the auroral oval and to the inner edge of the plasma

sheet, J. Geophys. Res., 87, 9059, !982.

(Ao= 66Ü) astrong pitch angle diffusion would lead
to a precipitation time scale of m20 minutes for

Hultqvist, B., et al., Quiet time convection electric
field properties derived from keV electron mea-

! keV electrons [Kennel, 1969]. Under such conditions, the SEB can be likened to a convection
boundary continuously fed by plasma sheet particles;
in fact,
scale can

surements at the inner edge of the plasma sheet by meansof GEOS-2, Planet. SpaceSci.,3__0,261,!982.
Jorjio, N.V., et al., Auroral diffuse zone. Part III:
Comparison between the equatorward boundary of the diffuse auroral zone and the plasmapause during the magnetic storms of Feburary !3 and17, 1972. Cosmic Res., 16, 750, !978. Kamide, Y., and J.D. Winningham, A statistical study of the instantaneous night-side auroral oval : The equatorward boundary of electron precipitation as observed by the ISIS-! and 2 satellites., J. Geophys. Res., 82, 5573, 1977. Kennel, C.F., Consequence of magnetospheric plasma,

the convection
be estimated

flow
to be

characteristic
about 2.5 hours

time
at

L = 6 for tosphere is weak, the SEB

a total potential drop across the magneof 30kV. The corresponding precipitation with a nearly empty loss-cone so'· that at the measured electron integral energy flux

can be as low as 3 x!0-3erg (cm2. sec.ster)-!. As
shownby Vondrak and Rich[1982] in such quiet time
periods an abrupt termination of the electric field and field aligned currents is observed at the equatorward boundary of the diffuse aurora; this is probably due to the development of an Alfv·n layer that significantly shields a further penetration
of the convection electric field earthward from

Rev. Geophys. Space Phys., 7, 379, 1969.
Khalipov, V.L., et al., Auroral diffuse zone. Part II: Formation and dynamics of the polar cliff of the subauroral ionospheric trough in the evening
sector, Cosmic Res., 15, 6!9, 1977. Kohnlein, W., and W.J. Raitt, Position of the midlatitude trough in the topside ionosphere as de-

the SEB[Southwood and Wolf, !978].
For periods of increasing magnetic activity, the inertia of the boundary can be accounted for by the fact that the polarization electric field breaks the equatorward displacement of SEB so that this motion can only be achieved by several sequential steps (each for a new substorm) followed by

duced from ESRO-4observations, Planet. Space
Sci., 25, 600, 1977.

Lui, A.T.Y., and C.D. Anger, A uniform belt of diffuse auroral emission seen by the ISIS-2 scan-

particle drift, precipitation, tivity enhancementand finally

ionospheric conducdischarge of the

ning photometer, Planet. SpaceSci., 2_[1,809, 1973. Lyons, L.R., Electron diffusion driven by magneto-

polarization charges. Once injectedtoward Earth, the
as relative plasmaclouds into dipolar field lines, a weak precipitation enables the particles to drift
for hours near the equatorial plane, as evidenced by the SEB memory during decreasing magnetic activity periods. This is in full agreement with the results deduced from low energy particle measure-

sphericelectrostatic waves,J. Geophys. Res.,
7__9, 575, !974.
Meng,C.I., et al., Electron precipitation of evening
diffuse aurora and its conjugate near the magnetospheric equator,
Res., Slater, 84, 2545, 1979. D.W., et al., Correlated

electron fluxes J. Geophys.
of

observations

ments at geostationary orbit [De Forest and McIlwain, 1971]. It must finally be stressed that
these results imply that aurora during disturbed variable low energy part belt located between the the source of the diffuse periods constitutes a of the outer radiation plasmapause [Jorjio et

the equatorial

diffuse

auroral

boundary,

J. Geophys. Res., 85, 53!,
Southwood, D.J., and R.A. the role of precipitation
Starkov,
Valchuk,

!980.

Wolf, An assessment of in magnetospheric
Auroral sub-

convection, J. Ge0phY.S. Res., 83, 5227, !978.
G.V.,
T.E.,

and Y.I.
et al.,

Felstein,

al., !978;Horwitz et al., !982] and the inner boundary of the region continuously fed by plasma of solar wind origin, i.e. the plasma sheet inner
boundary.
References

storm, Geomagn.Aeron., 3, 560, !97!.
Auroral diffuse zone. Part IV:

Latitudinal distribution of auroral optical emissions and particle precipitation and its relationship with the plasma sheet and magnetotail, Cosmic Res., !7, 459, !979.

Arnoldy, and Chan, tLL., K.W. Particlesubstorms obs- Vondrak, S3-2andF. R.R., satellite J·-·ich,Simultaneous Chatanika radar and measurements of
erved at the geostationary orbit.
74, 50!9, !969.

J. Geop.hy s.

Res.,

ionospheric

electrodynamics

in the diffuse

Belmont, G., et al., Are equatorial electron cyclotron waves responsible for diffuse auroral elec-

aurora, J. Geophys. Res., 87, 6!73,

!982.

tron precipitation Res., !983.
De Forest, S.E.,

? to appear, J..GeophY.s L.
(Received April 6, 1983;
Plasma clouds accepted ·ay 10, 1983.)

and C.E. McIlwain,