Nigh%me
auroras
for
remote
sensing
of
magnetospheric
processes
Anita
Aikio
University
of
Oulu
Finland
Photo:
J.
Jussila
Foreword
The
task
is
like
“Explain
the
universe
and
give
two
examples”.
=>
As
said
in
the
abstract,
I
will
try
to
cover
SOME
aspects
of
nigh%me
auroras
and
their
origin.
…
Luckily,
Eric
Donovan
has
almost
the
same
topic,
so
he
will
probably
explain
some
of
the
missing
parts
(e.g.
magnetospheric
substorm
processes).
Contents
•  Large‐scale
configura?on
of
aurora
•  Diffuse
aurora
•  Discrete
aurora
–  Electrosta?c
accelera?on
–  Alfvénic
aurora
–  Some
possible
genera?on
mechanisms
of
auroral
arcs
–  Seasonal
effects
–  North‐south
aurora
and
plasma
bubbles
•  Summary
Auroral
zone
•  Regions
in
which
night‐
?me
auroras
occur
in
the
sta?s?cal
sense.
•  Within
the
auroral
zones,
auroras
are
observed
by
the
naked
eye
on
more
than
50%
of
nights
(if
there
are
no
clouds),
even
during
years
of
low
solar
ac?vity.
•  They
are
centered
roughly
23o
from
the
geomagne?c
poles
and
are
~
10o
wide.
Figure
from
Kaila
(1998).
Auroral
zone
in
the
southern
hemisphere
•  Regions
in
which
night‐?me
auroras
occur
in
the
sta?s?cal
sense.
•  Within
the
auroral
zones,
auroras
are
observed
by
the
naked
eye
on
more
than
50%
of
nights
(if
there
are
no
clouds),
even
during
years
of
low
solar
ac?vity.
•  They
are
centered
roughly
23o
from
the
geomagne?c
poles
and
are
~
10o
wide.
Figure
from
Kaila
(1998).
Auroral
oval
Based
on
observa?ons
during
the
Interna?onal
Geophysical
Year
1956‐57,
the
instantaneous
distribu?on
of
auroral
ac?vity
versus
magne?c
local
?me
(MLT)
and
magne?c
la?tude
(MLAT)
was
found
to
be
an
oval‐shaped
belt
called
the
auroral
oval
by
Feldstein
and
Starkov
in
1967.
The
auroral
ovals
(one
for
each
hemisphere)
are
con?nuous
bands
centered
near
67o MLAT
at
magne?c
midnight
and
near
about
77o MLAT
at
magne?c
noon.
!"#$%&$!'$(')*+,'%$)*'("-$').%/%)'/0)*'1/%$2)!("&#'3'!0)*(%)24'""'
Figure
from
Sigernes
et
al.
(2010).
As
magne?c
ac?vity
increases,
the
oval
expands.
"#$%!&% !"#$%&'()%*+,+%),-%./)%/)%)0*"1&#,"),0 !") #"('2)34567)%"()&#$')0,+)89)&:) ;'1'$<'+)844=)
Auroral
oval
and
diffuse
aurora
!"#$%&$!'$(')*+,'%$)*'("-$').%/%)'/0)*'1/%$2)!("&#'3'!0)*(%)24'""'
Originally,
the
oval
was
parametrized
by
the
Q
index,
but
modern
versions
use
Kp
or
AL
index.
Equatorward
of
the
oval,
diffuse
aurora
exists
and
it
is
also
included
in
the
sta?s?cal
models.
"#$%! &%! "#$%&'$()**! +(',,-! ./012! 345! ,6/$78'9$'.! :8/-.$';! 8<! 7=,! .><</+,! $/'8'$?! 3@5!
$/'8'$! 8#$%! "#AB?! 3B5! 0$C-,7>(! -8'7=! 18%,?! 3D5! <>,%.! 8<! #>,9! $/'8'$! 8:+,'#,'?! 3E5!
8:+,'#,'! %8($7>8-?! 3F5! G88-! $-.! "/-! >-<8'0$7>8-! $7! %8($%! +>7,?! 3H5! 18+>7>8-! 8<! 7=,!
I$1>.,;,!4!+$7,%%>7,J &>0,2 KL2KK!M&!8-!@D!7=! N,(,0:,'!@KKLJ!
&=,!7'$-+<8'0$7>8-!78!C,8C'$1=>( (88'.>-$7,+!>+!7=,Figure
from
Sigernes
et
al.
(2010).
'%! ()*(+,-./0,1!2+,34"*+5!
&=,!8#$%+!($%(/%$7,.!>-!7=,!1',#>8/+!+,(7>8-!$',!(,-7',.!
8-! 7=,! 0$C-,7>(! 18%,+J! &=,! O$'7,+>$-! (8018-,-7+! >-!
7=,!C'$1=!$',!',%$7,.!78!7=,!18%$'!0$C-,7>(!(88'.>-$7,+!
:;
!'!
 '!$! 
T!  
 &! = )!
  &!$!  ?!
 
3F5!
Diffuse
and
discrete
aurora
18,662
Discrete
aurora:
•  structured
(e.g.
arc)
•  bright
emissions
Diffuse
aurora:
•  no
(observable)
internal
structure
•  rela?vely
low
luminosity
•  spread
fairly
uniformly
over
a
wide
area,
typically
in
the
equatorward
part
of
the
oval
•  RIGHT:
Sta?s?cal
loca?ons
of
diffuse
and
discrete
aurora
as
magne?c
ac?vity
increases
from
1
to
5.
The
lowest
la?tude
shown
is
60o
MLAT
(So?relis
and
Newell,
2000).
SOTIRELIS AND NEWELL: BOUNDARY-ORIENTED ELECTRON PRECIPITATION MODEL
(a)
#1.••
(b)
•,•- ..,:.........:::.::::::
....
:::.........:....
,•,,,• :g•.
••..'•½•;
..................
•:•.•,,.•:•...
•gf,-'".•=•:•.: .......
.
.::::::....:•:•:,..
...:::.
....
':':•
(c) ½3
(e) #5
Diffuse aurora
Structuredprecipitation
Subvisual drizzle
Figure2. Theaverage
locations
of thevarious
auroral
regions
for thefivedifferent
activity
levelsof the
model.Noonis at thetopwithduskat theleft.Concentric
circlesaredrawnat 60ø, 70ø, and80ø MLAT.
The
auroral
intensiQes
within
the
oval
SOTIRELIS
AND NEWELL:
BOUNDARY-ORIENTED
ELECTRON
(a) #•
(b) #2
(c) #3
(d) #4
PRECIPITATION
MODEL
(e) #5
UP:
Instantantaneous
satellite
measurement
of
auroral
intensi?es
in
the
oval
(false
colours).
RIGHT:
Sta?s?cal
energy
fluxes
of
electrons
by
satellite
measurements
as
the
magne?c
ac?vity
increases
from
1
to
5
(So?relis
and
Newell,
2000).
Median Spectra'
Energy Flux
13
12
11
10
9
".•' 8
Log JE
[eV/s-sr-cm2]
Plate4. Thelogof thedirectional
energyfluxJE(eV/ssrcm2)of electrons
in themedian
model
18,663
Magnetosphere
(3D
view)
Plasma
sheet
contains
plasma
that
originally
comes
from
the
solar
wind
or
from
the
ionosphere.
It
contains
the
most
energe?c
plasma
(electron
energies
typically
600
eV)
in
the
magnetosphere.
Plasma
sheet
boundary
layer
(PSBL)
is
the
outer
region
of
the
plasma
sheet
with
electrons
of
typically
150
eV
and
field‐aligned
electron
or
ion
beams
are
ohen
found
there.
IMF
Bz<0
(southward)
=>
idealized
steady
reconnecQon
+
plasma
convecQon
(Simula?on
by
Lockwood)
Energy
to
drive
magnetospheric
plasma
convecQon,
field‐aligned
currents
and
aurora
comes
from
the
solar
wind
A
part
of
the
energy
carried
with
the
solar
wind
to
penetrate
into
the
magnetosphere,
especially
when
IMF
has
a
southward
component.
One
widely
used
es?mate
for
the
energy
input
is
the
epsilon
parameter
by
Akasofu:
ε =
107
vB2sin4(θ/2)lo2
where
the
unit
of
ε
is
W,
lo
is
a
length
of
7
RE,
v
and
B
are
the
solar
wind
speed
and
magne?c
field,
respec?vely,
and
θ
is
the
clock
angle
of
the
IMF
defined
as
θ =
arctan(By/Bz).
Several
other
coupling
func?ons
exist
and
these
have
been
compared
to
different
indices
of
geomagne?c
ac?vity,
especially
to
AE,
AL
and
Dst
(e.g.
Newell
et
al,
2007).
Magnetospheric
substorms
When
IMF
Bz<0,
some
energy
is
stored
as
magne?c
energy
in
the
nightside
magnetotail.
The
explosive
release
of
this
energy
is
called
magnetospheric
substorm.
This
anima?on
shows
forma?on
of
the
near‐Earth
neutral
line
(NENL)
in
the
tail.
(Simula?on
by
Lockwood)
NEWELL ET AL.: PRECIPITATION MAP BINNED BY IMF
A10206
Origin
of
precipitaQon
by
DMSP
satellites
Bz<0,
By>0
Bz<0,
By<0
Newell
et
al.
(2004):
A10206
NEWELL ET AL.: PRECIPITATION MAP BINNED BY IMF
A10206
Nightside
oval
is
mainly
CPS
(central
plasma
sheet)
and
BPS
(boundary
plasma
sheet)
precipita?on.
CPS:
all
electron
spectra
are
Figure 1. Maps of ionospheric precipitation according to the source region, with superpose
Bz>0,
By<0
0, B > 0; (c) B > 0, B < 0; (d) B > 0
inertial convection streamlines: (a) for B < 0 and B < 0; (b) B <Bz>0,
By>0
similar,
“unstructured”.
B > 0.
BPS:
High
variability,
“structured”.
These
low‐al?tude
(835
km)
5 of 20
CPS
and
BPS
regions
are
NOT
necessarily
the
same
as
PS
and
PSBL
in
the
magnetosphere.
z
y
y
Figure 1. (continued)
z
y
z
y
z
The diffuse aurora, unlike the discrete aurora, is more difficult to see from the
Earth because it is relatively dim and lacks sharp outlines. However, it is much more
Auroral
forms
within
the
oval
expansive and spread out than the discrete aurora. See Figure 2.3 for the some of the
During
the
Interna?onal
Geophysical
Year
(IGY)
(1957‐58),
space
scien?sts
all
around
the
world
coordinated
their
efforts
to
record
the
aurora
from
many
places
at
the
same
?me.
More
than
100
all‐
sky
cameras
were
set
up
to
watch
the
aurora.
From
the
analysis
of
this
data,
two
important
concepts
in
auroral
physics
were
born:
"auroral
oval"
and
"auroral
substorm”.
One
of
the
persons
to
make
Figure 2.3. Shapes and Locations of the Discrete and Diffuse Aurora (Akasofu, 1981)
measurements
was
S.‐I.
Schema?c
figure
of
aurora
by
Akasofu.
Akasofu.
2-4
The diffuse aurora, unlike the discrete aurora, is more difficult to see from the
Earth because it is relatively dim and lacks sharp outlines. However, it is much more
Auroral
forms
within
the
oval
expansive and spread out than the discrete aurora. See Figure 2.3 for the some of the
•  Diffuse
aurora
equatorward
of
the
main
oval.
•  Discrete
auroral
arcs
•  Homogenous
bands
•  Curtains
or
draperies
•  Rayed
arcs
•  Folds
and
curls
•  North‐south
auroral
arcs
(streamers)
•  Polar
cap
arcs
(theta
aurora)
•  Omega
bands
•  Westward
travelling
surge
and
auroral
bulge
•  Auroral
patches
•  Pulsa?ng
or
flickering
Figure 2.3. Shapes and Locations of the Discrete and Diffuse Aurora (Akasofu, 1981)
aurora
•  Cusp
aurora
Schema?c
figure
of
aurora
by
Akasofu.
2-4
•  Black
aurora
(usually
stripes
or
arcs)
Origin
of
diffuse
aurora
•  It
has
been
believed
that
diffuse
aurora
is
caused
by
pitch‐angle
scarering
of
CPS
electrons
(0.1–30‐keV
)
into
the
loss
cone,
but
the
precise
mechanism
has
been
unclear.
•  Two
classes
of
magnetospheric
plasma
waves,
electrosta?c
electron
cyclotron
harmonic
(ECH)
waves
and
whistler‐mode
chorus
waves,
could
be
responsible
for
wave‐par?cle
interac?ons
that
lead
to
P‐A
scarering:
ω
–
kIIvII
=
nΩec
/
γ ,
where
ω is
the
wave
frequency,
Doppler
shihed
to
a
mul?ple
(n
=0,±1,±2,...)
of
the
rela?vis?c
electron
gyrofrequency
Ωec
•  Thorne
et
al.
reported
in
Nature
in
2010
that
sca\ering
by
whistler
mode
chorus
waves
(f
<
fec)
is
the
dominant
cause
of
the
most
intense
diffuse
auroral
precipita?on.
ETALB=.IGHTSIDECHORUSISSTRONGLYCONFINEDTOTHE
EQUATORIALREGION<15WHILEDAYSIDEEMISSIONSARESTRON
GERATHIGHLATITUDES>20!SIGNIFICANTCORRELATIONHAS
BEENFOUNDBETWEENTHESTORMTIMEACCELERATIONOFELECTRONS
TORELATIVISTICENERGIESTHROUGHOUTTHEENTIREOUTERRADIATION
BELTSANDENHANCEDCHORUSEMISSIONS;-EREDITHETAL
C=ORMICROBURSTPRECIPITATION;/"RIENETAL=
ec
SUGGESTINGTHATCHORUSPLAYSANIMPORTANTROLEINTHEACCEL
ERATIONPROCESS
Origin
of
diffuse
aurora
Whistler
mode
chorus
waves
(f
<
f )
recorded
on
the
ground
(leh)
and
expected
distribu?on
in
space
in
a
recovery
phase
of
a
storm
(right).
DEN
DRA
DUR
WA
TIC
-E
!
DRI
THE
THE
PLA
MU
FREQ
(courtesy
of
J.
Manninen)
&IGURE 3CHEMATICMODELFORTHEEXPECTEDDISTRIBUTIONOFPLASMA
WAVESDURINGTHERECOVERYPHASEOFTHELARGE(ALLOWEEN
(Thorne,
2005)
STORM4HEPLASMAPAUSEPOSITIONWASOBTAINEDFROM)-!'%
OBSERVATIONSCOURTESY*'OLDSTEIN
WH
AND
AND
MA
CAN
TO
OF
CAL
Energy
of
parQcles
in
discrete
aurora
Energy
of
plasma
sheet
and
PSBL
electrons
is
typically
smaller
than
for
electrons
producing
aurora
in
the
ionosphere.
Further
energy
may
come
from:
•  Auroral
accelera?on
region
at
al?tudes
of
~1
–
3
RE
•  During
substorms,
from
magne?c
reconnec?on
in
the
tail
at
NENL
(X~20‐30
RE)
•  From
waves,
especially
from
kine?c/iner?al
Alfvén
waves
generator
region
aurorae
aurorae
low-B
high-B
AcceleraQon
of
auroral
parQcles
by
a
quasi‐staQc
electric
field
structure
(potenQal
drop)
Satellites
above
auroral
arcs
ohen
measure
electron
energy
spectra
that
are
referred
to
as
“inverted‐V”
events.
In
those,
the
maximum
characteris?c
energy
occurs
in
the
center
of
an
auroral
arc.
That
has
been
interpreted
to
be
produced
by
a
quasi‐sta?c
electric
field
structure
and
an
associated
“parallel
poten?al
drop”,
where
precipita?ng
electrons
gain
a
maximum
energy
of
eVII.
Electron
energy
between
1
and
10
keV
shown
by
dashed
lines.
(Auroral
Plasma
Physics,
2002)
AcceleraQon
of
auroral
parQcles
by
a
quasi‐staQc
electric
field
structure
(potenQal
drop)
EI
EI
Satellites
above
auroral
arcs
o_en
measure
electron
energy
spectra
that
are
referred
to
as
“inverted‐V”
events.
In
those,
the
maximum
characterisQc
energy
occurs
in
the
center
of
an
auroral
arc.
That
has
been
interpreted
to
be
produced
by
a
quasi‐staQc
electric
field
EII
structure
and
an
associated
“parallel
potenQal
drop”,
where
precipitaQng
electrons
gain
a
maximum
energy
of
eVII.
EI
Electron
energy
between
1
and
10
keV
shown
by
dashed
lines.
p+
e­
AcceleraQon
of
auroral
parQcles
by
a
quasi‐staQc
electric
field
structure
(potenQal
drop)
1986-05-02 Viking o384
Satellite
measurements
at
al?tudes
of
1‐3
RE
(the
accelera?on
region)
have
confirmed
the
picture
by
electric
field,
par?cle
and
magne?c
field
measurements,
e.g.
Mozer
et
al.
(1980),
Temerin
et
al.
(1981),
Weimer
(1985),
Marklund
et
al.
(1997),
Carlson
et
al.
(1998),
Marklund
et
al.
(2001),
…
300
EI
NS
E2bsp (mV/m)
200
100
0
-100
-200
0
VII
P2bsp (kV)
-1
-2
-3
-4
-5
0
bEW
b3bsp (nT)
-20
-40
-60
-80
jup
jup (micro A/m^2)
-100
2
1
0
-1
-2
51:00
51:30
52:00
Time 15 UT + minutes
52:30
Aikio
et
al.
(1996)
AcceleraQon
of
auroral
parQcles
by
inerQal
Alfvén
waves
ENS (mV/m)
bEW (nT)
FAST
satellite
measurement
of
wave
fields
and
electrons
close
to
the
nightside
polar
cap
boundary
at
an
al?tude
of
about
2740
km
showing
evidence
of
iner?al
Alfvén
waves
and
F‐A
electrons
accelerated
by
the
waves.
Note
that
electron
energy
range
is
broad.
e– energy (eV)
e– p-a (deg)
Energy flux (erg cm-2s-1)
(Auroral
Plasma
Physics,
2002)
Current
system
associated
with
auroral
arcs
(Marklund
suggested
that
the
return
current
could
be
associated
with
“black
aurora”
in
some
cases)
Figure 7.3: Schematic figure of the auroral current circuit with negative and positive
potential structures corresponding to convergent and divergent high-altitude electric
fields, respectively (Marklund, 2001).
SchemaQcs
from
high‐alQtude
satellite
measurements
Carlson
et
al.
(1998)
Widths
of
auroral
arcs
N. Partamies et al.: Observations of auroral width spectrum
Partamies
et
al.
(2010)
200
Knudsen
et
al.
(2001)
180
Maggs
and
Davis
(1968)
160
Number of events
140
120
100
80
60
40
20
0
0.1
0.1
1
1
Width (km)
Width
(km)
10
10
Fig. 8. Histograms of the scale sizes of the auroral structures. The
black one is from the statistics of Maggs and Davis (1968), the red
Some
theoreQcal
consideraQons
related
to
observed
widths
of
auroral
arcs
ElectrostaQc
coupling
between
the
magnetosphere
and
the
Figure 6.4: The field-aligned current density away from the ionosphere (upwar
ionosphere:
function of the parallel potential drop for various values of the mirror ratio Bi /B
•  Natural
scale
length
L
of
electrosta?c
mapping
between
the
linear regime is clearly visible (Auroral Plasma Physics, 2002).
MAGNETOSPHERE-IONOSPHERE
COUPLING
77
ionosphere
and
magnetosphere
(Weimer,
1985)
depends
on
Pedersen
conductance
Σ
exercise) and then
P
and
field‐line
conductance
K
!
2
ne
ΣP
and
K
is
j! =
1 V! = KV! ,
L=
(6.21)
(2πme Eth ) 2
K
tic scale length, which
is alsoKtermed
the as
magnetosphere-ionosphere
where
is known
the Lyons-Evans-Lundin constant or the field-aligned condu
It
has
been
es?mated
that
L
varies
between
100
km
(e.g.
ngth. The numeric value
of this
depends on the
well the energy flux and in the limit k
From
the constant
same assumptions,
onerather
can derive
Lysak,
1991)
and
340
km
(Borovsky,
1993)
=>
large‐scale
Bi
typically
from 5 to 50 S and the much
P , which for auroral arcs
eVvaries
! ! Bm kB T one gets
inverted‐V’s
could
be
explained.
value of K, which has
been estimated to range from 10−11 S m−2 to
2
ten quoted value for L is 100 km, which is obtained by using ne
values
ε=
K = 1 · 10−9 S m−2 .
we get
(2πme Eth )
1
2
V!2 = KV!2 .
Some
theoreQcal
consideraQons
related
to
observed
widths
of
auroral
arcs
How
quasi‐staQc
EII
can
be
produced
at
accelera?on
region
al?tudes?
Sugges?ons
include:
•  Anisotropy
between
electron
and
ion
pitch‐angle
distribu?ons
(so
that
ions
mirror
at
lower
al?tude)
•  Strong
plasma
double
layers
(VII>kBTe/me)
•  Large
number
of
weak
double
layers
(VII<kBTe/me)
•  Electrosta?c
shocks
•  Anomalous
resis?vity
(by
turbulence
and
plasma
waves)
The
minimum
width
of
aurora
produced
by
the
mechanisms
above
was
es?mated
by
Borovsky
(1993)
to
be
1‐2
km
mapped
to
ionospheric
(100
km)
al?tudes.
However,
according
to
observa?ons,
more
typical
values
are
some
tens
of
km.
Some
theoreQcal
consideraQons
related
to
CHAPTER 6. MAGNETOSPHERE-IONOSPHERE
COUPLING
observed
widths
of
auroral
arcs
96
which is larger
than for shear Alfvén waves (cf. eq. (6.61)) and the parallel electric field
The
parallel
electric
field
in
inerQal
Alfvén
wave
is
given
by:
of the wave
is given by
E!
k! k⊥ λ2e
=
.
(6.92)
2 2
E⊥
1 + k ⊥ λe
where
λ
and
k
are
parallel
This parallel
electric
field accelerates the cold electronsIIto
carry the field-aligned current.
e
is
the
electron
skin
depth
and
k
and
perpendicular
wavenumbers.
Typically
λek
~1.
EII
can
be
The FAST
satellite observations indicate that typically k⊥ λe ∼ 1. The magnitude
1
mV/m
over
wavelengths
of
more
than
1000
km.
The
of E! canpoten?al
drop
in
the
wavefront
of
the
iner?al
Alfvén
wave
can
be 1 mV/m over wavelengths of more than 1000 km. Consequently the
potential drop in the wavefront of the inertial Alfvén wave can be of the order of kilovolts,
be
of
the
order
of
kilovolts,
though
more
typically
hundreds
of
though more typically hundreds of eV. A cool electron can be picked up by the wave
eV.
Typical
widths
are
~1
km.
and accelerated by the wave parallel electric field towards the ionosphere. Fig. 6.15
shows an example of FAST data, which shows correlation of wave Poynting flux and
the downgoing (pitch angle 0◦ , second panel from bottom) electron beam energy. The
son, 1985;R. Benson,privatecommunication,
1991]or at fieldlinesmap to the daysideauroralzone(cleftregion).
highatltitudes[Bahnsenet al., 1989].Because
the thick- Note that the mappingof the magnetopause
is difficult to
nessof an auroralarc in this modeldependsso stronglyon
discernin Plate 4 of Kaufmann ½t al. and so the valueof
Some
theoreQcal
consideraQons
related
to
observed
widths
of
auroral
arcs
the parameters
of the lower-hybrid-wave
turbulence
(satu- œm•g/œion
hasa largeuncertainty
associated
with it. Using
rationamplitudes
andparticleheatingrates),andbecause thesevaluesof Bmag/Bion
andœmag/œion
in expression
(2)
theseparametersare as yet completelyuncertain,no arc- yieldsWion-- Wm•g/76
- 130km. Thisvalueisentered
into
thicknesspredictioncan be made with this model. This is
Table 1.
reflected in Table 1.
Velocity
shears
(plasma
power(current
and voltage)to auroral-arc
magnetic-field
vor?city
given
by
x
v)
can
act
lines are investigatedand an auroral-arcthicknessvalueis
as
a
MHD
generator
and
drive
obtained
for eachmechanism.Thesevaluesare compiled
into Table1. The 10 generatormechanisms
are thought
F‐A
currents.
Borovsky
(1993)
to operatein variouslocationsin the magnetosphere
and
ionosphere,
andsothe mappings
of thesemechanisms
to the
es?mated
that
typical
shear
ionosphere
will bemoreinvolved
thanwerethe mappings
of
the acceleratormechanisms
of section3, whichall had the
widths
in
in
the
low‐la?tude
same location.
boundary
layer
(LLBL)
give
d.1. Shearin the Low-LatitudeBoundaryLayer
widths
of
~130
km
and
in
the
The inertia
associated
with a plasmaflowacross
a magneticfieldcan act as an MHD generator,drivingelectrical
currentsplasma
sheet
3
–
50
km.
from regionswhere there is a shearin the cross4. GENERATOR MECHANISMS
In this section10 mechanisms
that cansupplyelectrical
fieldvelocity
[Schmidt,
1960;Rosa,1961;Borovsky,
1987a].
There is some belief that the Earth's low-latitude
Zones
of
Strong Shear
d•
dx•
Zones
of
Strong
Shear
bound-
ary layer mapsto the auroral zone and that shearedflows
in the low-latitude boundary layer drive auroral currents
[Eastmanet al., 1976;Sonnerup,1980;Bythrowet al.,
1981; Heikkila, 1984; Lundin and Evans, 1985; Lotko et
al., 1987;Phan et al., 1989; Siscoeet al., 1991;Lundin
et al., 1991].Forinertialflowsof plasmaacross
magnetic Fig. 8. A sketchof a broad regionof velocityshear with two
fields,the magnitudesof the currentdensitiesdrivenare re-
zonesof more-intense
sheaxwithin (seesection4.1).
vation ∇ · j! = −∇ · j⊥ gives
!
"
Some
theoreQcal
consideraQons
related
to
B × ∇p
∇ · j! = −∇ · j⊥ = −∇ ·
.
(5.64)
2
B
observed
widths
of
auroral
arcs
CHAPTER 5. PLASMA CONVECTION AND MAGNETOSPHERIC CURREN
Pressure
gradients
in
the
plasma
sheet
may
be
associated
with
FAC:
n (Vasyliunas,
1970; Heinemann and Pontius, 1990) that this equation
#ion j
eq
Beq
= − 2 · ∇peq × ∇V ,
B
Beq
!
(5.65)
asyliunas equation. Here V is the differential flux tube volume (i.e. the
gnetic flux tube of unit magnetic flux). This volume is given by
$ ion
ds
V =
,
(5.66)
B
eq
Borovsky
(1993)
es?mated
that
for
4
keV
ions
in
the
plasma
Figure 5.15: Example of contours of constant plasma pressure (red) and fl
sheet
(X=
‐12
Re)
the
minimum
pressure
gradient
width
gral is extended
along a magnetic field line from the equatorial plane to
volume (blue) in the equatorial plane, which give rise to a field-aligned curren
(corresponding
to
ion
gyroradius)
would
map
to
a
value
of
~3
km
If, for simplicity, ionosphere
we assume
that jPlasma
in 2002).
the equatorial plane,
! vanishes
(Auroral
Physics,
in
the
ionosphere.
s the parallel
current density in the ionosphere. This approach doesn’t
ration mechanism, Current
it just conservation
addresses diversion
from
perpendicular to
∇ · j! = −∇
· j⊥the
gives
Some
theoreQcal
consideraQons
related
to
observed
widths
of
auroral
arcs
Theories
related
to
ionospheric
conducQvity
feedback
instability
(e.g.
Sato
and
Holzer,
1973;
Sato
1978;
Lysak
1991)
have
the
following
elements:
•  Ini?ally
a
small
perturba?on
(increase)
in
ionospheric
conductance
exists
•  In
the
presence
of
a
convec?on
electric
field,
polariza?on
EF
is
created
due
to
ENS
component
•  The
perturba?on
EF
travels
to
the
magnetosphere
with
Alfvén
speed
and
is
associated
with
up‐
and
downward
FACs
at
wave
front
•  Due
to
EEW
component,
perturba?on
in
the
ionosphere
drihs
in
the
NS‐
direc?on
(ExB
–drih)
•  When
the
perturba?on
reaches
the
source
region
in
the
plasma
sheet,
electrons
are
drawn
toward
the
ionosphere
•  Due
to
ionospheric
plasma
convec?on,
electrons
may
(or
may
not)
hit
the
original
conduc?vity
enhancement
and
further
increase
conduc?vity
•  Borovsky
(1993):
widths
of
17
–
83
km.
Some
theoreQcal
consideraQons
related
to
observed
widths
of
auroral
arcs
Conclusion:
Several
mechanisms
produce
auroral
arcs
with
widths
above
1
km,
but
small‐scale
(<
1
km)
widths
are
difficult
to
explain!
Seasonal
effects
on
energeQc
electron
precipitaQon
Newell
et
al.
published
in
1996
in
Nature
an
ar?cle:
“Suppression
of
discrete
aurorae
by
sunlight”
They
found
that:
•  the
beams
of
accelerated
electrons
that
cause
intense
discrete
aurorae
occur
mainly
in
darkness:
the
winter
hemisphere
is
favoured
over
the
summer
hemisphere,
and
night
is
favoured
over
day
(by
a
factor
of
3)
•  discrete
aurora
rarely
occur
in
the
presence
of
diffuse
aurora
Also,
other
phenomena
related
to
electrosta?c
accelera?on
show
the
same
seasonal
varia?on:
•  intense
electric
fields
in
auroral
accelera?on
region
(Marklund
et
al.,
1994)
•  upflowing
ion
beams
(Collis
et
al.,
1998)
•  auroral
kilometric
radia?on
(Kumamota
and
Oya,
1998)
Seasonal
effects
on
energeQc
electron
precipitaQon
Probability
of
observing
accelerated
e–
(monoenerge?c)
aurora
(for
energy
fluxes
>
5
erg/cm2s‐1)
Darkness
Sunlight
Newell
et
al.
(1996)
Seasonal
effects
on
auroral
parQcle
precipitaQon
Conclusion
by
Newell
et
al.
(1996):
Ionospheric
conducQvity
plays
a
role,
e.g.
by
the
ionospheric
feedback
instability.
Another
explana?on
suggested
is
that
forma?on
of
parallel
electric
field
requires
low
background
densiQes
at
high
alQtudes
(density
cavi?es),
which
MIGHT
form
preferen?ally
when
ionospheric
electron
densi?es
are
low
in
the
en?re
field
line.
Seasonal
effects
on
broadband
(wave)
aurora
A03216
NEWELL ET AL.: SEASONAL AURORAL VARIATIONS
Winter
Summer
A03216
Newell
et
al.
(2010):
Wave
(Alfvenic)
aurora
occurs
in
the
region
of
typical
substorms.
It
also
has
larger
energy
fluxes
in
darkness
than
in
sunlight,
but
the
difference
is
not
as
large
as
for
monoenerge?c
electrons.
Possible
explana?on:
interhemispheric
currents
(Lyatskaya
et
al.,
2008).
Seasonal
effects
on
broadband
(wave)
aurora
L20104
LYATSKAYA ET AL.: SUBSTORMS AND INTERHEM
Sugges?on
for
interhemispheric
currents
by
Lyatskaya
et
al.
(2008):
Upward
FAC
in
the
pre‐midnight
sector
(electron
precipita?on)
Figure 3. Schematical view of currents during northern
Part
of
Region
I
currents
from
the
summer
hemisphere
is
diverted
to
the
summer. The currents of substorm current wedge are shown
winter
hemisphere,
where
field‐aligned
and
horizontal
currents
flow
in
the
in blue, the Region 1 currents are shown in red. Interhemispheric field-aligned currents (IHC) are branched from the
same
direc?on
as
substorm
current
wedge
currents.
Note:
upward
FAC
Region 1 currents, flow along the magnetic field to opposite
coincides
with
electron
precipita?on
and
aurora.
(winter) hemisphere and close the ionospheric currents in
the southern auroral zone.
Christiansen, F.,
tions of highMagsat obser
2001JA90010
Engebretson, M
ULF wave pow
southern days
Boundary Lay
and T. Onsage
Fujii, R., T. Iijim
dence of larg
1103 – 1106.
Huang, C.-S. (20
changes and m
doi:10.1029/20
Hurtaud, Y., C. P
and diurnal ef
plasma conve
A09217, doi:1
Liou, K., P. T. N
particle accele
5542.
Lyatsky, W., an
of geomagne
2007SW00038
Lyatsky, W., A.
electrojet fro
doi:10.1029/20
Lyatsky, W., S. L
North‐south
aurora
and
plasma
bubbles
Bursty
bulk
flows
(BBFs)
in
the
magnetotail
are
transient
high‐
speed
(vx≥
400
km/s)
flows
which
contain
depleted
(low
Ne)
flux
tubes.
Sergeev
et
al..
(2000)
North‐south
aurora
and
plasma
bubbles
Bursty
bulk
flows
(BBFs)
in
the
magnetotail
are
transient
high‐
speed
(vx≥
400
km/s)
flows
which
contain
depleted
(low
Ne)
flux
tubes.
Nakamura
et
al..
(2001)
The
main
explana?on
for
BBFs
is
the
plasma
bubble
model
by
Pon?us
and
Wolf
(1990):
The
bubble
moves
earthward
due
to
Epol
(interchange
instability),
and
is
associated
with
F‐A
currents.
North‐south
aurora
and
plasma
bubbles
North‐south
aurora
was
suggested
by
Henderson
et
al.
(1998)
to
be
asociated
with
bursty
bulk
flows
(BBFs)
in
the
magnetotail.
The
bubble
model
is
the
most
plausible
model
to
explain
BBFs
and
observa?ons
confirm
that
the
upward
FAC
of
the
bubble
is
associated
with
N‐S
aurora
KAURISTIE ET AL.: BBF INTRUSION TO THE INNER PLASMA SHEET
(also
called
streamers)
(e.g.
Kauris?e
et
al.,
2003;
Pitkänen
et
al.,
2011).
Auroral
streamer
SIA
jup
Juusola
(2010).
(Kauris?e
et
al.,
2003)
Figure 1. Auroral streamers observed by the UV imager onboard the Polar satellite (top) and by the
ASC at MUO (bottom). The pink circle in the first plot of the top panel show the field of view of the
Summary
Ground‐based
op?cal
observa?ons
show
that
mul?ple
scale
structures
are
ohen
observed
in
aurora.
To
explain
those,
probably
several
different
genera?on
and
accelera?on
mechanisms
are
needed.
At
the
moment,
no
uniform
theory
neither
general
consensus
exists
=>
Key
challenge
for
magnetosphere‐ionosphere
coupling
studies.
!"#$%&
4
'$%&
5$6##$&
($)*$+,-,
($)*$.),/01&2/)3