Download Global convection electric field and current : Comparisons

Global convection electric field and current :
Comparisons between model’s predictions and data from
STARE, SAINT-SANTIN and magnetometers
C. Mazaudier, C. Senior, E. Nielsen
To cite this version:
C. Mazaudier, C. Senior, E. Nielsen. Global convection electric field and current : Comparisons
between model’s predictions and data from STARE, SAINT-SANTIN and magnetometers.
Journal of Geophysical Research, American Geophysical Union, 1987, 92 (A6), pp.5991-5999.
HAL Id: hal-00976880
http://hal.upmc.fr/hal-00976880
Submitted on 10 Apr 2014
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diffusion de documents
scientifiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
JOURNAL
OF GEOPHYSICAL
Global
RESEARCH,
Convection
VOL. 92, NO. A6, PAGES 5991-5999, JUNE
Electric
Field
1, 1987
and Current:
Comparisons Between Model's Predictions and Data
From STARE, Saint-Santin, and Magnetometers
C. MAZAUDIER
AND C. SENIOR
Centre de Recherchesen Physique de l'Environnement,St.-Maur des Fosses,France
E. NIELSEN
Max-Planck lnstitut fiir Aeronomie,Katlenburg-Lindau, Federal Republic of Germany
Ionospheric electric field and ground magnetic field measurementsat high, middle and low latitudes
are used to study the global pattern of the convectionelectricfield on March 26, 1979. The IMF B:
component was southward and relatively steady over a prolonged period of time. The data set is thus
representativeof a steady magnetosphericconvection. A semi-analytical time-dependent convection
model is used to interpret the observations.During the daytime the model steady state results roughly
reproduces the mean observed convection pattern. Fluctuations of the observed electric field from the
model predictionsare interpretedas resultingfrom (1) the transientresponsesof the ionosphereto small
time variations of the IMF, (2) substorm effects,(3) neutral wind disturbances.Only the first type of
fluctuationscan be reproducedby the initial stateprediction of the convectionmodel.
1.
made by both coherentand incoherentradars [Brekke et al.,
INTRODUCTION
Axford and Hines [1961] and Dungey [1961] first suggested
that solar wind/magnetosphereviscousinteraction and magnetic reconnection transfer solar wind energy and momentum
to the magnetosphereand ionosphere.This coupling createsa
large-scalecirculation of the magnetosphericplasma (current
convection cells). Since then, it has gradually been recognized
that the ionospherealso plays an important role in determining this plasma flow. Different methods have been developed
to study the ionosphericelectriccurrent and field generatedby
the solar wind-magnetosphereinteractions.
First, the observations of the ground magnetic variations
1974; Zi and Nielsen, 1981, 1982; Mazaudier et al., 1984].
Third, theseobservationswere interpretedby usingtheoretical models which considered the effects on the electric field
distribution of ionosphericconductivities,ring current and interplanetary magneticfield (IMF). The influenceof the magnetosphericring current on the ionosphericelectric field was
investigated in analytical studies [Vasyliunas, 1970; 1972;
Pellat and Laval, 1972; Southwood and Wolf, 1977] semianalytical studies [Senior and Blanc, 1984], and numerical
methods [Jaggi and Wolf, 1973; Hard et al., 1981a, b; Spiro et
al., 1981; Wolf et al., 1982; Chen et al., 1982; Spiro and Wolf,
ledto theestablishment
of theSqpandDPe equivalent
current 1984].
systems[Na•lata and Kokubun, 1962; Nishida, 1968]. Both
current systemsare composedof two cells.They differ only in
their latitudinalextension:the Sq• currentsystemremains
To progressfurther in the experimentaldeterminationof
the global convection patterns, multi-instrument studies are
essential. We have made observations
of convection
electric
confined at high latitudes, while the DP e one extendsto low fieldsby meansof coherentand incoherentradars (Scandinalatitudes. Nishida was the first to interpret these current sys- vian twin auroral radar experiment(STARE) and Saint-Santin
tems as having their origin in the interaction betweenthe solar radars), and equivalent current systems(Troms/5 magnetometer), and comparedtheseobservationswith model predictions
wind and the magnetosphere.
in
order to understand the solar wind, magnetosphere,and
Second,observationsof the ionosphericelectric fields led to
ionosphere
interactions.
the recognitionthat the convectionelectricfieldsgeneratedby
In a companionpaper, Mazaudier et al. [1984] carried out
the solar wind/magnetospheredynamo penetrate to middle
and low latitudes. This is contrary to an earlier belief that the an analysisof this kind, using magnetic and radar observaonly ionosphericregion fed by the energy available at the tions together with convectionmodel predictions.This first
solar wind/magnetosphereinterfaceis the auroral zone. Inco- study investigatedthe transient responseof ionosphericelecherent scatter radars have played an essentialrole in the ob- tric fields to an increaseof magnetosphericconvectiondue to
servations of these fields at the equator [Fejer et al., 1979; a sudden southward turning of the IMF B: component.The
Gonzales et al., 1979; Kelley et al., 1979], at midlatitudes present paper is concernedwith steady state convection,i.e.,
[Wand and Evans, 1981; Blanc, 1983a, b] and at both lo- with the global convection pattern observed during a procations simultaneously[Gonzaleset al., 1983]. At high lati- longed period of southwarddirectedIMF B•. We presentobtudes, in the auroral oval, important contributions have been servations made during an 18-hour period during which the
Copyright 1987 by the American GeophysicalUnion.
Paper number 6A8600.
0148-0227/87/006A-8600505.00
IMF was southward and showedsmall magnitude fluctuations
in time. In the following section, we present the geomagnetic
conditions characterizing this time interval. We then relate
electricfield and magneticobservationsat high, middle, and
low latitudes to the Senior and Blanc [1984] model predic5991
5992
MAZAUDIER ET AL,' GLOBAL CONVECTION ELECTRIC FIELD AND CURRENT
0
6
12
18
2t, UT TIME
0
The total crosspolar cap potential drop •o and the local
time of the maximum potential at the polar cap boundary are
taken, respectively,as 70 kV and 0400 LT for the simulation.
These values are chosento fit the best the high latitude data
Bz (IMF)
from the STARE
DST
-S0
(b)
formularelatingthe IMF B• componentto •o givenby Re/ff
+500
AU
o
(c)
AL
-500
330
et al. [1985].
3.1. High Latitudes
H Component
(d)
(Y) !
Tromsi•
(69øN)
0
6
12
18
radar. The value of 70 kV is smaller than the
estimation of ~ 110 kV that is calculated from the empirical
2t• UT TIME
Fig. 1. The four panels outline the magnetic activity on March
26, 1979. (a) Bz componentof the IMF measuredby the IMP 8
satellite; (b) equatorial magnetic Dst index; (c) AU and AL auroral
magnetic indices; (d) H component of the ground magnetic field at
Troms6.
Figure 2 showsa polar plot of the E x B drift as measured
by the STARE coherent radar system,[Nielsen and Whitehead, 1983; Nielsen and Schlegel,1983, 1985]. It illustratesthe
typicalconvectionpattern at high latitude,with westwardion
drift in the afternoon sector and eastward in the morning
sector. Two reversals of the drift occur around
tions. This linear time-dependentmodel estimateselectricfield
due to the sole physicalprocessof direct penetrationof magnetosphericconvection.
2.
GEOMAGNETICCONDITIONSON MARCH 26, 1979
March 26, 1979,is a magneticallydisturbedday (ZKp - 31,
ZKn- 27). The amplitude of the IMF as measured by the
IMP-8 satellite (located near the magnetopausein the solar
wind) is quasi steadyduring the entire day (around 20 nT); the
2100 and 2200
MLT (1930 and 2030 UT). Around 1200 MLT (1030 UT), the
amplitudeof the auroral electricfield was probably smaller
than 15 mV/m, the thresholdat which the STARE radar can
measure electric fields. The convection pattern observedon
March 26, 1979, correspondsto the average convectionpattern obtainedby Zi and Nielsen[1982].
In Figure 3 we have superimposedthe STARE southward
electric field measured at the latitude
of 69øN extracted from
the Figure 2 and the B• componentof the interplanetarymagnetic field. There is some correlation
between the maxima
of
Bycomponent
remainspositive(around18 nT). The Bz com- the electricfield and the southward intensificationof the B•
ponent, shown in the first panel of Figure 1, is negative centered around
the mean value of -3
nT from 0000 to 1800 UT.
After 1800 UT, its magnitude decreasesand Bz turns northward at about
2230 UT.
The Dst index (Figure lb) shows two major negative excursions at 1200 and
nT.
The
1800
A U and AL
UT.
auroral
This
index
indices
remains
are shown
below
-25
in the third
panel. These curvesreveal sustainedactivity during the entire
day with two clear substorm onsetsshown with arrows. The
most intense substorm begins around 1700 UT with a maximum AL of about 1200 y at 1800 UT. At 2230 UT, when the
IMF turns northward, the auroral activity stops.In the European sector, the Troms6 H component (Figure ld) remains
undisturbed from 0500 to 1100 UT, while the Dst is quasi
steady.From 0000 to 0500 UT, the Troms6 station is under
the westward electrojet, and from 1200 to 1700 UT under the
component.This suggeststhat variationsof the B• component
of the IMF is an important parameter governing the ionosphericelectricfields.
On Figure 4 we superimposedthe diurnal variation of the
H componentof the ground-magneticperturbationfor March
26, 1979, at Troms/5(samedata as Figure ld) and the averaged
Sd variation of H over Troms/5 obtained by Mayaud [1965].
The averaged Sd variation is the mean value of the five most
disturbeddays (following the Kp index) of each month. This
Sd curve givesthe magneticvariation that would be observed
by a station if it were rotating under a systemof two convection current cells of fixed intensity and location. Hence the
65
eastward one. At the time of the 1700 UT substorm, the
Troms6
H trace shows westward
directed currents for a short
period, before being again dominated by the eastward electrojet. From 2030 to 2230 UT Troms6 moves back into the
westward electrojet. At 2200 UT, the auroral activity at
Troms6
ends.
3.
OBSERVATIONS AND RESULTS
In this section, we present separately the IMF measurements in the solar wind, and the electric field at high, middle
and low latitudes and we compare some of these data to the
outputs of the convection model of Senior and Blanc [1984].
The logic and main assumptionsof this model are describedin
the appendix. Briefly, it calculates the time-dependent responseof the ionosphere to a step function of the convection
source,i.e., the crosspolar cap potential •o. The electricfield
pattern has been shown to reach a steady state after 30-60
min. Thus for steady interplanetary conditionssuch as those
prevailing on March 26, 1979, we compare the data to the
model outputs at steadystate.
18I[
6
ß •;'_'-
"-< 5:
0 (MLT)
22.30 (UT)
Fig. 2. Polar plot of the ion drift observed at high latitudes by the
STARE radar, in an MLT-invariant latitude coordinate system.
MAZAUDIER
ET AL.' GLOBAL CONVECTION ELECTRIC FIELD AND CURRENT
5993
uT
Fig. 3. Southwardcomponentof the STARE electricfield and B: componentof the IMF. The southward/northward
electricfieldintensifications
in theauroralzonearegenerallyassociated
with southward
increases
of
differencebetweenthe two curveson Figure 4 can be interpreted as due to variations in intensity and/or location of the
current cells. The best agreement between the two curves is
obtained on the dayside between 0500 and 1800 UT. It shows
that during that time interval, the European sector moves
under a steady convection pattern. Before 0700 UT and after
1700 UT the H component at Troms/3 fluctuates and reaches
maxima which correspondin generalto the averagedSd value
except during the main substormaround 1800 UT.
According to Chapman and Bartels [1940] the variation of
the ground magnetic field is related to the height-integrated
ionosphericcurrent densityby
IABI
=
f. JI
12"
(la)
for an infinite current sheetabove a plane earth.
In expression(la), B is in 7, J is in amperesper kilometer,
andf is equal to 0.6 [Kamide and Brekke, 1975]. An horizontal
ionosphericeastwardJy currentinducesa variationin the
horizontal component of the magnetic field
AH _• Jy
(lb)
Following Ohm's law, and assuminghomogeneousconductivity [see Wilkinsonet al., 1986] the eastwardcurrent density
is related to the ionosphericelectricfield by
J• -• Z.E x
(2)
where Z n is the height-integratedHall conductivity,E•, is the
northward electricfield in the plane perpendicularto the magnetic field. From expressions(lb) and (2) we can derive
AH -• Zn,,
(3)
Figure 5 illustrates the relation between the ground magnetic perturbation and the southward electric field. In this
figure the STARE southward electric field measured in the
region over Troms/3 is superimposedon the H component at
106
I
I
I
_h,•
.
i
I% I
I •/
I
II
I
0 1 2 3/ Ir•A 5/ 6.'' 7 8 9 10 11 12 13 16 '5 16 17 t',8
q',19' 20 •1 22
•
f
:.....
(-...,
•
I
I I II
uT.Time
,
Fig. 4. The H componentof the ground magnetic field obserwd at Troms6 on March 26, 1979 (solid li•), is
superimposed
on the mean Savalue of the H componentestimatedby Mayaud [1965] (dashedline).This Sd variationis
the mea• valueowr a solarcycleof the fiw most magneticallydisturbeddaysof eachmonth.
5994
MAZAUDIER ET AL.: GLOBAL CONVECTION ELECTRIC FIELD AND CURRENT
Fig. 5. The southwardSTARE electricfield and ground magneticH componentat Troms• are superimposed.
A good
correlation is found between the southward/northward intensificationsof the electric field and the northward/southward
increasesof the magnetic field, exceptduring the 1800-2000 UT substormevent.
Troms/5.A very good agreementbetweenthe variations of AH ponent. Its amplitude has been chosen as 70 kV, with a maxiand E,, is found except from 1700 to 2000 UT, during the mum at 0400 LT.
substorm event. The ratio between the H component and the
The auroral conductivityhas beenselectedby comparing
southward electric field gives an estimate for the height- the Troms/5H componentand the northward STARE electric
integrated Hall ionospheric conductivity in the auroral zone. field.In Figure6a the modelsouthwardelectricfield is superFor this data set the Hall conductivity varies between 5 and imposedon our STARE data. The modelreproducesthe mean
12 mhos. These two extremas
have been selected for simula-
tions.
diurnal variation observedexceptduring the 1800 UT substorm event. The model obviously does not reproduce the
The convection model of Senior and Blanc [1984], which fluctuations
of the electricfieldrelatedto the Bz fluctuations.
considersthe direct penetration of convection to midlatitudes This is because we consider the convection electric field to be
(seethe appendix), has been used to determine the steady state inducedby a steadypotential drop over the polar cap. The
ionospheric electric field. The two simulations developed in agreement between model predictions and data for the eastthis study are describedin Table 1. Some of the input parame- ward componentof the electricfieldis ratherpoor (Figure6b).
ters have been derived from our observations.
The electric
This is in fact not very surprisingsincethis componentis
potential •0 along the polar cap/auroralzone boundaryhas knownto be veryweakand to showlittle systematic
magnetic
been adjusted to fit the STARE northward electric field com- activity dependenceI-Seniorand Blanc, 1984; Mazaudier et al.,
TABLE 1. Inputs of the ConvectionModel for the Two
Simulations
S 1 and S2
S1
Latitudinal Range
S2
CrossPolar Cap Potential
j. = 72 ø
•o = 70 kV*
•o = 70 kV*
Conductivities
Auroral
zone
67 ø _< J. _< 72 ø
Subauroral
zone
64 ø _< 2 _< 67 ø
Midlatitudes
10 o _</• _< 64 ø
Z n = 12 mhos
Zn/Ze = 2
Ea = 5 mhos
ZH/Z P '-' 2
mhos
1.6
Z n - 5 mhos
Y•n/Y•v= 1.6
daytimeZn = Z•ø(Bo/B)cosX
nighttime•n = (•nø/(30)(Bo/B)
•n/•v = 1.7
Equatorial region
2 _< 10 o
equatorialnoontimeconductance
SO
So=3.83 x 108f•m
same local time dependence
as midlatitude
*Maximum
at 0400 LT.
conductivities
same as S 1
same as S 1
MAZAUDIER
ET AL.' GLOBAL CONVECTION ELECTRIC FIELD AND CURRENT
60
1• •o:
70kV
Hmax
=0400LT
Stare
(X
=68
ø)
Simulation
E
5995
/
I I-"•-Stare
data
----
,
,
- 20-
,
1
Simulation
2
,
,
..
.
,
•
model.
•
40
•
-60
-
Fig. 6a. The southwardcomponentof the STARE electricfield on March 26, 1979 (solid line), is superimposedon the
southward convection electric field obtained from two distincts numerical simulations of the Senior and Blanc [1984]
model. The model predictionscorrespondto the steady state of magnetosphericconvection.
1984]. It is interesting to note that the model steady state
convectionelectric field (Figure 6a) is fairly well related to the
Sd variation of the H componentat Tromsb (Figure 4, dashed
line). The Sd variations,like the electricfield model one, are
nisms by case studies. Mazaudier et al. [1985] developed an
analysis of dynamo region neutral wind disturbancesat SaintSantin generated by storm Joule heating. These papers
smoothed.
tion is frequently observed on the dayside at midlatitudes. By
contrast the ionospheric disturbance dynamo processesdevelop less frequently on the dayside in the midlatitude dynamo
region.
On March 26, 1979, the neutral wind analysis was performed at Saint-Santin in the F region during the night, and in
the E region during the day, when the incoherent scatter
signal was strong enough. The data reveal strong neutral wind
disturbances at F region altitudes on the morningside from
0000 to 0700 UT. Remembering these points we can comment
on Figures 7a and 7b. On these figures we superimpose the
southward (Figure 7a) and eastward (Figure 7b) electric field
model and Saint-Santin data. The data represent the disturbed
electric field obtained by subtracting from the observations
the magneticallyquiet-time electric field model as defined by
Blanc and Amayenc [1979]. On the dayside from 0700 UT to
1700 UT (local time and universal time are the same at Saint-
showedthat the direct penetrationof magnetosphericconvec-
The steadystate model reproducesthe $d amplitude of the
H componentat Tromsb if we assumean enhancedHall conductivity auroral ring of 10-12 mhos, uniform in longitude
within
the auroral
zone. This indicates
that
simulation
2 is
closer to the actual data than simulation 1 (seeTable 1).
3.2.
Middle
Latitudes
At middle latitudes, electric field disturbancesare produced
by various physical mechanisms.Two of thesemechanismsare
particularly important: (1) the penetration of convection electric fields to subauroral latitudes during periods of slow
change in the solar wind dynamo (our interest in this paper);
(2) changesin global thermosphericcirculation produced by
storm-time Joule heating affecting the ionospheric wind
dynamo [Blanc and Richmond,1980].
Mazaudier [1985] has recently illustrated this two mechaq'o= 70kV
Stare {X =68• )
Hmox: 0400LT
-----
Simulation
1
S•mulation
2
E 20
___•_Steady
states
ofthe
convection
model
• o
•
• •i
I
i
I
I
I II
I
•
I
•:20 •
_
Fig. 6b. Similar to Figure 3b for the eastwardcomponentof the auroral electricfield.
5996
MAZAUDIER ET AL.' GLOBAL CONVECTION ELECTRIC FIELD AND CURRENT
% = 70 kV
Hmax= 0400 LT
Saint-Santin( X =/,/.,*)
Simulation
1
-----Simulation
2 .
Santin
data
/
I
i
•9 /
••
I
,
I
I
[
i
•
,
,
,
•
I J [
J • / LT.
hours
........ --,
Ls
ready st
•o= 70kV
Saint- Santin ( X = 44* )
Hmax=0400LT
----•
Simulation
1
Simulation
2
2
Fig. 7. Our two convectionsteadystatemodelpredictionsare superimposed
on the (a) southwardand (b) eastward
components
of the disturbedelectricfield at Saint-Santin.The disturbedfield is obtainedby subtractingthe quiet time
variation definedby Blancand Arnayenc[1979] from the observations.
Santin), the steady state convection model give a rough estimate
of the
southward
electric
field
variation
observed
on
sentsthe usual equatorial electrojet of the quiet ionospheric
dynamo.Following Akasofuand Chapman[1961] the ground
magneticvariation can be expressedby
March 26, 1979. The same agreement is obtained for the eastward component (Figure 7b) on the dayside.It should be noAH = DR + DP + DCF + D T
(4a)
ticed however that, for midlatitudes, the convection model
predicts similar convectionpatterns on the dayside at the ini- in this expressionDR, DP, DCF, and D T correspondto ring,
tial and steady states. On the nightside strong discrepancies ionospheric,magnetopause
and tail currents,respectively.
are observed between model and data for both components.
On both March 21 and 26, due to the fact that there is no
Those occurring on the morningside (from 0000 to 0700 UT)
magnetic storm, the equatorial variation reduced to [Fukucan be attributed
to neutral wind disturbances.
Those ocshimaand Kamide, 1973]
curring on the eveningside(after 1700 UT) can be explained
AH = DR + DP
(4b)
by, the substorm that affects the Troms6 area from 1800 to
2000 UT, and by the northward turning of Bz around 2230 The symmetricpart of the DR field is approximatedby Dst
UT, both effects that are not taken into account in the model.
cos 2, where Dst is the magnetic equatorial index and 2 the
latitude
3.3.
Equatorial Latitudes
of the station.
To emphasize the DP disturbance on March 26, 1979, it is
Figure 8a is composed of two panels, the upper one for
Addis-Ababa (geographic latitude 9.03øN, geographic lon-
necessaryto have a quiet referencefor the ionosphericDP
current system.This referenceis representedby the circled
gitude 38ø77E, LT -• UT + 3) and the bottom one for Huan- crosseson Figure 8a. It is obtainedby subtractingthe DR field
cayo (12.05øS,284ø67E, LT • UT- 5). On these two panels (DR -- Dst cos 2) from the H componentof the ground magwe have superimposed the H component of the terrestrial netic variations of the quiet referenceday (March 21). Theremagnetic field on March 26, 1979 (dashed line), and the same fore, the differencebetweenthe H componentof ground magcomponent observed on March 21, 1979; March 21 was a
netic field on March
magneticallyquiet day, that we use as a reference.It repre-
the ionosphericDP disturbanceon that day.
26 and the circled crosses is a measure of
MAZAUDIER ET AL..' GLOBAL CONVECTION ELECTRIC FIELD AND CURRENT
90J;[
3
6
FHorch,
21
1979
quiet reference doy
9
12
5o•
[
0
]
5997
15
ADDIS-
18
ABABA
.-
21 .........
CQ•_
._•D_.
?_-•,
2•,
UT.Time
•
3 LT.Time
HUANCAYO
,,-";•
,,'•
6
9
LT
"ø
UT
5
•. ..............
12
./J',;
•
'• © x('5•xxrch
21,1979
lB
_!,._A
•©
•)._
. )1-•.•
V. UT. Time
12 - 1/+
",,"
'•"•.....1•.
LTTime
North
26, 1979-1
Fig. 8a. Ground magnetic H component on March 26, 1979 (dashed lines), and March 21, 1979 (solid line) at
Addis-Ababa (upper panel) and Huancayo (lower panel). March 21, 1979, is a magnetically quiet day used as a reference
level. On the two panelsthe crossesrepresentthe quiet day observationscorrectedfor the Dst variations on March 21 and
26, 1979.
At Addis-Ababa, the equatorial DP disturbance on the dayside observed on March 26, 1979, is southward, and corresponds to a westward equivalent current. It occurs from 1100
to 1500 LT (0800 to 1200 UT). At Huancayo the equatorial
DP disturbancefluctuates and showsstrong eastward intensifications of the electrojet from 0700 to 1300 LT (1200 to 1800
UT).
In Figure 8b the northward component of the Stare electric
of neutral wind disturbanceson the nightside until 0700 UT.
It is thus possible that these disturbances propagate southeastward toward
It is clear that
the northward
intensifications
of the auroral
southwardincreasesof Bz and are the signatureof the DP2
equivalent current system.
The convection model of Senior and Blanc [1984] which
includesa rough description of the electrojet region through
its low-latitude boundary condition (see appendix and Table
1), predicts the total equatorial currents at initial state and
steady state (see Figure 9). On the dayside, at the initial state,
the total electrojet increaseis eastward. This variation is similar to the H component fluctuations observed at Huancayo.
The total current in the equatorial electrojet was estimated
from the Huancayo data by assumingthe jet to be 10ø latitude
wide in the northern hemisphereas in the model, and to have
uniform electric current density of 100 A/km (100 nT from
Figure 8b). The total current in the equatorial electrojet is
then 100 kA; this order of magnitude correspondsto the initial state model predictions.
At Addis-Ababa (Figure 8a upper panel), the daytime electrojet disturbance is westward with an intensity of about 100
kA. This is not in good agreementwith the predictions of the
convectionmodel. It is thereforeprobablethat another mechanism operateslocally at Addis-Ababa. In fact, Blanc and Richmond [1980] have shown that the ionospheric disturbance
dynamo has a negativeeffect on the equatorial electrojet,ie a
westward increaseof the current during the daytime. Study of
the F region neutral winds at Saint-Santin reveal the presence
with a time constant of several
hours, characteristicof the ionosphericdynamo process.The
result would be westward electrojet disturbanceas observedat
Addis-Ababa. Huancayo which is located west of Saint-Santin,
would not seetheseperturbations.
4.
field and the DP disturbanceat Huancayo are superimposed.
zone electric field correspondto the eastward equatorial electrojet intensifications.These fluctuationsare also associatedto
Addis-Ababa
DISCUSSION AND CONCLUSION
In this paper we have combined magnetic data with coherent and incoherentradar data, and comparethesedata to the
predictions of a convection model. This enables us to analyze
quantitatively the large-scaleconvection electric field pattern
on March 26, 1979.
On this particular day, (1) the IMF Bz component was
southward and rel.atively steady over a prolonged period of
time from 0000 to 1900 UT, (2) there was no neutral wind
disturbance at midlatitudes in the dynamo region from 0700
to 1800 UT. This made it possible to study the prolonged
action of the magnetosphericconvectionprocessfrom 0700 to
1800 UT. The convection model of Senior and Blanc [1984]
was used to calculate the ionosphericelectric field. The steady
state model prediction was used to establishthe mean diurnal
variation
of the electric field due to the mechanism
of direct
penetration of magnetosphericconvection.
12
15
,
electrojet
d•s •
v
18
UT
"
IJ
•electric
•/-•---northword
store
field
mc•rch26,1979
Fig. 8b. Comparisonbetweenthe DP intensificationof the equatorial electrojet and the northward STARE electric field.
5998
MAZAUDIERET AL.' GLOBALCONVECTIONELECTRICFIELD AND CURRENT
EqUQfOrIQI perfurbQfion
APPENDIX'
DESCRIPTION OF THE
SENIORAND BLANC [1984] MODEL
100
/ •
- SfeQdy
sfQfe
-•
• so i/I \\\
This linear time-dependentand self-consistentmodel calcu-
lates the responseof the magnetosphere
ionospherecircuit to
a stepfunctionof the source,i.e., the crosspolar cap potential.
It takesinto accountthe latitudinal and longitudinalgradients
of the midlatitudeionosphericconductivitiescreatedby solar
radiations.
O0
Or,
08
12
16
Locol Time_
hours
20
Fig. 9. Localtime response
of the total equatorialelectrojetcurrent predictedby the convectionmodelfor the initial (dashedline)
and steady state (solid line) of magnetosphericconvectionevent
[Seniorand Blanc, 1984,Figure 17].
This model keeps the simplicity (and also indeed the limitations) of its linearity and has freely adjustablevaluesof the
high latitude conductivities.The conductivityin the auroral
zone is simulatedwith two uniform conductivityrings' the
auroral zone from colatitudes0o to 02 and a subauroralzone
from 02 to 0•. For 0 greaterthan 0• (middleand low latitudes)
the integratedPedersenconductivityvariesduring the daytime as
This paper showsthat on the daysidethe steadystateof the
Senior and Blanc convectionmodel givesa rough estimateof
the ionosphericconvectionpattern in the latitude range from
the auroral
zone to midlatitude.
Bo
Zv= Zvøcos
Z-•where Z is the solar zenith angle, B is the local value of the
magnetic
field,B0 and•,o arethe valuesof theequatorial
At auroral latitudes, the main diurnal variation of the magnetic field and conductivity at local noon. During the
southwardSTARE electricfield is roughly reproducedby the night the midlatitude conductivities are small and uniform in
steadystatemodelas well as the Sd variationof the H compo- local time'
nent observedat Tromsb [Mayaud, 1965].
E•øB0
E•-30 B
The southward/northward observed electric field intensi-
fications,usuallyassociated
with southwardBz increases,
are
not reproducedby the steadystatemodel.They wereinterpreted as due to transient responsesof the ionosphereto the
southward Bz fluctuationsand reproducedby initial state of
The subsolar Pedersen conductance has been taken as 20
northward
time variations as the midlatitude
Mhos and a smoothtransition betweenthe day and night
values,modeledas a parabolicfunctionof Z, is imposedon the
the convection model.
dawn and dusksectorsover 20ø of the parameter.Finally, the
At middle latitudes,the daytime variations of the two elec- midlatitude Hall to Pedersenconductivityratio is 1.7 and the
tric field componentsare well reproducedby the convection magnetic field is dipolaf with an equatorial value of 0.28
x 10 -'• T.
model, both the initial and steady statesof the model (which
The equatorial electrojetregion is included in the model as
are closeto each other on the dayside).
At Huancayo, closeto the equator, the H componentof the a boundarycondition.It is a small conductingring though
ground magneticfield showedstrongeastwardintensifications which the meridional currentsclose.The equatorial conducof the equatorial electrojet associatedwith increasesof the tance is 3.83 x 108 mho m at noon and has the same local
STARE
auroral
electric field. These observations
conductivities.
The inner boundary of the equatorial ring current is as-
were not reproducedby the steadystate model, which predicts
a small westward electrojet intensification associated with
northward electric field penetration as due to the transient
sentsthe equatorial boundary of the auroral zone. The model
response
of the ionosphere
(initial statemodel)to the IMF B•
yields self-consistent
variationsof the inner boundaryin the
sumedto map dawnon the circleof colatitude0• whichrepre-
fluctuations(•o crosspolar cap potential). The discrepancies magnetosphericequatorial plane and of the LT distribution of
Birkelandcurrentsgeneratedalongthis boundary.The colatibetween convection model and data at Addis-Ababa has been
tudes
0o and 02 and conductivitiesin each zone are free painterpreted as a local effectdue to the ionosphericdisturbance
rametersof the model and can be adjustedto fit the observadynamo process.
A previouscompanionpaper [Mazaudier et al., 1984] illustrated the initial state of convectionelectricfield inducedby
rapid changein the solar wind magnetosphere
dynamo (southward turningsof B0. This paper showedthat even during a
tions.
Acknowledgments.
The STARE radar systemis operatedby the
Max-Planck-Institut ffir Aeronomiein cooperationwith ELAB, the
Universityof Trondhum,Norway,and the FinnishMeteorological
time of prolongedsouthwardIMF B•, as on March 26, 1979, Institute in Helsinki, Finland. The extensionof the CNET (Centre
incoherentscatterfacility
it is difficultto find a perfectsteadystateof ionosphericcon- Nationald'EtudesdesT616communications)
vection electricfield in the whole ionosphere.Rather, the data
can be analyzed in terms of a mean diurnal variation correspondingto the steadystate model, and of short-termfluctu-
ationsproducedby transientchangesof the sourcepotential
due to IMF
variations.
It seems therefore that a linear model
to a quadristatic configuration was supported by the Institut
d'Astronomieet de G6ophysiqueand by the Direction desRecherches
et Moyens d'Essais.The facility is operatedwith financial support
fromtheCentreNationaldela Recherche
Scientifique.
The Editor thanks M. Harel and J. M. Holt for their assistancein
evaluating this paper.
of the convectionprocessis sufficientto interpret quantitatively the electric field observationswhen there is no substorm
or neutral wind disturbanceoperating simultaneouslyto the
convection process.
REFERENCES
Akasofu,S.-I.,andS.Chapman,A neutrallinedischarge
theoryof the
aurora Polaris,Philos.Trans.R. Soc.London,Set. A, 253, 359-406,
1961.
MAZAUDIER ET AL.: GLOBAL CONVECTION ELECTRIC FIELD AND CURRENT
Axford, W. I., and C. O. Hines, A unifying theory of high latitudes
geophysicalphenomenaand geomagneticstorms, Can. J. Phys., 39,
1433, 1961.
Blanc, M., Magnetospheric convection effects at midlatitudes, 1,
Saint-Santin observations,J. Geophys.Res.,88, 211, 1983a.
Blanc, M., Magnetospheric convection effectsat midlatitudes, 3, Theoretical deviation of the disturbance convection in the plasmasphere,J. Geophys.Res.,88, 235, 1983b.
Blanc, M., and P. Amayenc, Seasonal variations of the ionospheric
E x B drifts, J. Geophys.Res.,84, 2691, 1979.
Blanc, M., and A. Richmond, The ionosphericdisturbance dynamo, J.
Geophys.Res.,85, 1669, 1980.
Brekke, A., J. R. Doupnik, and P.M.
Banks, Incoherent scatter
measurementsof E region conductivities and currents in the auroral zone, J. Geophys.Res., 79, 3773, 1974.
Chapman, S., and J. Bartels, Geomagnetism,
Oxford University Press,
New York, 1940.
Chen, C. K., R. A. Wolf, M. Harel, and J. L. Karty, Theoretical
magnetogramsbased on quantitative simulation of magnetic substorm, J. Geophys.Res.,87, 6137, 1982.
Dungey, T. W., Interplanetary magnetic field and the auroral zones,
Phys. Rev. Lett., 6, 47, 1961.
Fejer, B. G., C. A. Gonzales, D. T. Farley, M. C. Kelley, and R. F.
Woodman, Equatorial electric field during magnetically disturbed
conditions, 1, The effect of the interplanetary magnetic fields, J.
Geophys.Res., 84, 5797, 1979.
Fukushima, N., and Y. Kamide, Partial ring current models for
worldwide geomagnetic disturbances, Rev. Geophys., 11, 4, 795,
1973.
Gonzales, G. A., M. C. Kelley, G. C. Fejer, J. F. Vickrey, and R. F.
Woodman, Equatorial electric fields during magnetically disturbed
conditions,2, Implications of simultaneousauroral and the equatorial measurements,J. Geophys.Res.,84, 5803-5812, 1979.
Gonzales, C. A., M. C. Kelley, R. A. Behnke, J. F. Vickrey, R. Wand,
and J. Holt, On the latitudinal variations of the ionosphericelectric
field during magnetospheredisturbances,J. Geophys.Res.,88, 9135,
1983.
Harel, M., R. A. Wolf, P. H. Reiff, R. W. Spiro, W. J. Burks, F. J.
Rich, and M. Smiddy, Quantitative simulation of a magnetospheric
substorm, 1. Model logic and overview, J. Geophys.Res., 86, 2217,
1981a.
Harel, M., R. A. Wolf, R. W. Spiro, P. H. Reiff, C. K. Chen, W. J.
Burke, F. J. Rich, and M. Smiddy, Quantitative simulation of a
magnetospheric substorm, 2. Comparison with observations, J.
Geophys.Res.,86, 2242, 1981b.
Jaggi, R. K., and R. A. Wolf, Self-consistentcalculation of the motion
of a sheet of ions in the magnetosphere,J. Geophys.Res., 78, 2852,
1973.
Kamide, Y., and A. Brekke, Auroral electrojet current density deduced from
Chatanika
radar
and from the Alaska
meridian
chain
of magneticobservatories,J. Geophys.Res.,80, 587, 1975.
Kelley, M. C., B. G. Fejer, and C. A. Gonzales, An explanation for
anomalous equatorial ionospheric electric fields associatedwith a
northward turning of the interplanetary magnetic field, Geophys.
Res. Lett., 6, 301, 1979.
Mayaud, P.M., A propos de la variation journalibre Sq du champ
magnbtique terrestre par la variation journalibre SD et d'un type
special de perturbations contribuant au SD d'•t•, Ann. Geophys.,
21, 219, 1965.
Mazaudier, C., Electric currents above Saint-Santin, 3. A preliminary
study of disturbance, June 6, 1978; March 22, 1979; March 23,
1979, J. Geophys.Res.,90, 1355, 1985.
Mazaudier, C., M. Blanc, E. Nielsen,and M.-Y. Zi, Latitudinal profile
of the magnetosphericconvectionelectric field at ionosphericaltitudes from a chain of magneticand radar data, J. Geophys.Res.,89,
375, 1984.
Mazaudier, C., R. Bernard, and S. V. Venkateswaran, Saint-Santin
radar observationsof lower thermosphericstorms,J. Geophys.Res.,
90, 2885, 1985.
$999
Nagata, T., and S. Kokubun, An additional geomagneticdaily vari-
ationfield(s•v field)in the polarregionon geomagnetically
quiet
days, Rep. Ionos. Space. Res. Jpn., 16, 256, 1962.
Nielsen, E., and K. Schlegel, A first comparison of STARE and
EISCAT electrondrift velocitymeasurements,J. Geophys.Res.,88,
5745, 1983.
Nielsen, E., and K. Schlegel, Coherent radar doppler measurements
and their relationship to the ionospheric electron drift velocity, J.
Geophys.Res., 90, 3498, 1985.
Nielsen, E., and J. D. Whitehead, Radar auroral observations and
ionosphericelectricfields,Adv. SpaceRes.,2(7), 131-144, 1983.
Nishida,A., A coherenceof geomagnetic
DP 2 fluctuationswith interplanetary magneticfield variations,J. Geophys.Res., 73, 5549, 1968.
Pellat, R., and G. Laval, Remarks on the steady and time mathematical convection models, in Critical Problems of Magnetospheric
Physics,edited by E. R. Dyer, Inter-Union Commission on Solar
Terrestrial Physics,Washington, D.C., 1972.
Reiff, P. H., R. W. Spiro, R. A. Wolf, Y. Kamide, and J. H. King,
Comparison of polar cap potential drops estimated from solar
wind and ground magnetometerdata: CDAW 6, J. Geophys.Res.,
90, 1318, 1985.
Senior, C., and M. Blanc, On the control of magnetosphericconvection by the spatial distribution of ionospheric conductivities, J.
Geophys.Res.,89, 261, 1984.
Southwood, D. J., and R. A. Wolf, The role of hot plasma in magnetosphereconvection,J. Geophys.Res.,82, 5512, 1977.
Spiro, R. W., and R. A. Wolf, Electrodynamicsof convection in the
inner magnetosphere, in Magnetospheric currents, Geophys.
Monogr. Ser., vol. 28, edited by T. A. Potemra, p. 247, AGU, Washington, D.C., 1984.
Spiro, R. W., M. Harel, R. A. Wolf, and P. H. Reiff, Quantitative
simulation of a magnetosphericsubstorm, 3, Plasmasphericelectric
fields and evolution of the plasmapause,J. Geophys.Res., 86, 2261,
1981.
Vasyliunas,V. M., Mathematical models of magnetosphericconvection and its coupling to the ionosphere,in Particles and Fields in
the Magnetosphere,edited by B. M. McCormac, p. 60, D. Reidel,
Hingham, Mass., 1970.
Vasyliunas, V. M., The interrelationship of magnetosphericprocess,in
Earth's Magnetospheric Processes,edited by B. M. M½Corma½, p.
29, D. Reidel, Hingham, Mass., 1972.
Wand, R. H., and J. V. Evans, The penetration of convection electric
fields to the latitude of Millstone Hill (2 = 56ø), J. Geophys.Res.,
86, 5809, 1981.
Wilkinson, P. J., E. Nielsen, and H. Lurh, Ionospheric conductances:
Associations between ionospheric E region electric fields and
ground magnetometervariations, J. Geophys.Res.,91, 5839, 1986.
Wolf, R. A., M. Harel, R. W. Spiro, G. H. Voigt, P. H. Reiff, and C.-K.
Chen, Computer simulation of inner magnetosphericdynamics for
the magnetic storm of July 29, 1977, J. Geophys.Res., 87, 5949,
1982.
Zi, M.-Y., and E. Nielsen, Spatial variations of ionospheric electric
fields at high latitudes on magnetic quiet days, in Exploration of the
Polar Upper Atmosphere,edited by C. S. Deehr and J. A. Holtet, p.
293, Hingham, Mass., 1981.
Zi, M.-Y., and E. Nielsen, Spatial variation of electric fields in the
high-latitude ionosphere,J. Geophys.Res.,87, 5202, 1982.
C. Mazaudier and C. Senior, Centre de Recherchesen Physique de
l'Environnement, 4 Avenue de Neptune, 94107 St.-Maur des Fosses
Cedex, France.
E. Nielsen, Max-Planck-Institut
f/Jr Aeronomie,
Postfach 2D,
D-3411 Katlenburg-Lindau, Federal Republic of Germany.
(ReceivedJune 16, 1986;
revised February 9, 1987;
acceptedFebruary 25, 1987.)