Transcript COSPAR WORKSHOP ON CAPACITY DEVELOPMENT

COSPAR WORKSHOP ON
CAPACITY DEVELOPMENT
May 2004 – Beijing, China:
SUBSTORMS!
Presentation by:
Robert L. McPherron
Institute of Geophysics and Planetary Physics and
Department of Earth and Space Science
University of California Los Angeles
May 6, 2004
Importance of Substorm Timing
•
•
•
•
This recent paper
concludes reconnection
can not be the cause of a
substorm!
A more appropriate
conclusion would have
been that the authors
believe that the start of the
expansion phase is not
due to reconnection
It is undeniable that
reconnection extracts
energy from the solar wind
and is required to release
it from the tail lobes
Timing analysis is not
simple but the event
described here suggests
that reconnection was the
likely cause of this
particular substorm onset
“Finally, a simple timing analysis of specific substorm
onset time features suggests that near-Earth
reconnection cannot occur before current disruption and
therefore cannot be the cause of substorm.”
The Auroral Oval from Space
• The aurora occurs in a narrow
band circling the magnetic pole
called the auroral oval
• The dayside is obvious from
UV dayglow on the left side of
the figure, midnight is at the
east coast of the US
• Aurora is caused by charged
particles hitting the
atmosphere and exciting
electrons to higher energy
levels which decay producing
light
DEFINITIONS OF A SUBSTORM
•
•
•
•
“A magnetic storm consists of sporadic and intermittent polar disturbances,
the lifetimes being usually one or more hours. These I call polar
substorms.” [Chapman, 1962]
“The sequence of auroral events over the entire polar region during the
passage from auroral quiet through the various active phases to subsequent
calm is called an auroral substorm: it coincides with a magnetic (DP)
substorm, with which it has some close relationships.” [Akasofu, 1964].
“To generalize the concept of the auroral substorm to include the worldwide
disturbance characteristics and to emphasize the importance of the
magnetosphere in auroral zone observations, we have suggested the term,
magnetospheric substorm.” [Coroniti, McPherron and Parks, 1968]
“A magnetospheric substorm is a transient process initiated on the night
side of the Earth in which a significant amount of energy derived from the
solar wind-magnetosphere interaction is deposited in the auroral ionosphere
and magnetosphere.” [Rostoker et al., 1980]
Magnetic Disturbances during a Polar
Magnetic Substorm
CANADA E-W CHAIN X COMPONENT (5-sec) 27-Aug-2001 03:00:00
•
•
•
A stack plot of N-S magnetic
field variations from east to
west across Canada
(midnight from 4 – 8 UT)
Positive bays (perturbations)
occur from dusk to midnight
Negative bays occur from
midnight to dawn
The disturbances begin
gradually and then are
suddenly intensified
1600
03:38:59 03:52:35
04:06:45 04:18:48
stj
1400
ott
1200
X Disturbance (nT)
•
pbq
gil
1000
rab
800
smi
600
sim
daw
400
mea
200
0
03:00
vic
03:15
03:30
03:45 04:00 04:15
Universal Time
04:30
04:45
05:00
Ionospheric Currents in a Polar Magnetic
Substorm
•
•
•
The polar magnetic substorm is caused by two current systems (DP-2 and DP-1)
DP-2 consists of two electrojets (east and west) flowing towards midnight
DP-1 is a current system centered at local midnight that flows within the region
of bright aurora
Satellite Image of Auroral Expansion
•
•
•
The aurora during a substorm usually begins as a small spot close to midnight
With time the aurora expands westward and poleward
The DP-1 current system flows through the region of bright aurora
Akasofu
Auroral Substorm
•
•
•
Akasofu used all sky
camera data to describe
the local time development
of aurora during a polar
substorm
He divided the substorm
into two phases: expansion
and recovery
He identified a number of
auroral features including:
–
–
–
–
–
Sudden brightening
Formation of westward
traveling surge
The auroral bulge
Eastward drifting omega
bands
Pulsation patches
Why is There a Tail?
Viscous Interaction
•
•
•
•
•
•
Dynamic pressure produces normal
stress on the magnetosphere
boundary and creates distorted field
lines that are closer to the Earth than
those of a dipole
The existence of stretched field line
requires tangential stress, or a force
parallel to the boundary
In ordinary fluid flow tangential stress
(or drag as sometimes called) is
caused by the viscous interaction
Particles are scattered from the
moving fluid into the stationary fluid
adding momentum, i.e. exerting a
force on the boundary
A flowing boundary layer is produced
In a collisionless plasma particle
scattering is not very efficient, but
surface waves and wave
transmission are also effective
MAGNETIC RECONNECTION
(Dungey, PRL, 6(2), 47, 1961)
•
•
•
•
•
Resistivity develops at subsolar point
Solar wind field merges with earth’s field
Solar wind transports flux over poles
Open field reconnects at distant x-line
Magnetic flux returns to day side
E  V  B
E
V
B
•
•
•
•
Open field lines continue in solar wind
Solar wind electric field is projected to
earth’s poles
Ionospheric plasma drifts in response
Pedersen and Hall currents flow in
ionosphere
SOLAR WIND TRANSPORT OF
DAYSIDE MAGNETIC FLUX
(Cowley, Rev. Geophys., 20(3), 531, 1982)
• Two fundamental processes
couple solar wind to
magnetosphere
• Reconnection at subsolar point
opens dipole field lines and
moves flux over the poles
• Viscous interaction moves
closed field lines along the
flanks
• Reconnection produces
stronger convection when IMF
is south
Ionospheric Projection of
Magnetospheric Convection
•
•
•
•
•
What happens at the feet of the field lines?
They also move in the same direction as the
equatorial crossing
The insulating atmosphere allows the
ionospheric field lines to “appear” to slip
relative to the remainder of the field line in the
neutral atmosphere and Earth
Thus field lines undergoing reconnection move
from the x-line at noon subsolar point, across
the polar cap to the nightside x-line where they
reconnect, and then return at low latitudes to
the dayside
Closed field lines in the low latitude boundary
layer (llbl) also move backward over the edges
of the polar cap and return at higher latitudes
IONOSPHERIC CONVECTION
•
•
The ionospheric plasma flows along equipotentials of the ionospheric
electric field
The flow is from noon to midnight across the polar cap then east and west
along the auroral oval back to the dayside
Flux Balance in Convection
•
•
•
•
•
It is important to consider the
reservoirs of magnetic flux that
exist in a dipole field
These include the dayside
magnetosphere (closed field
lines), the tail lobes (open field
lines), and the nightside plasma
sheet (closed field lines)
In general it is unlikely that the
rate of change of flux in all
regions are balanced
These rates depend on the
merging rate M, the transport rate
T, the reconnection rate R, and
the return rate R
A steady state tail occurs when all
rates are equal
Field Aligned Currents Related to
Magnetospheric Convection
•
•
•
•
•
•
•
The low latitude boundary layer moves tailward
just inside the magnetopause
The electric field within the dawn side layer points
outward
Just inside the boundary layer the plasma flows
Sunward so the electric field is inward
Magnetic field lines map the magnetospheric
electric fields to the ionosphere
On the dawn side the E field point poleward
across the polar cap and equatorward
The electric fields drive currents in the ionosphere
that diverge at the point where the inner edge of
the boundary layer maps
The current divergence is fed by a field-aligned
current called the Region 1 current
“WHAT CAUSES A SUBSTORM?”
IMF - Bz and RECONNECTION!
•
•
•
•
Use 30 years of hourly AE and IMF
Bz
For each 1 nT bin of Bz create the
cumulative probability distribution
for corresponding AE index
Display contour map of the
probability of observing AE
exceeding a given value as a
function of Bz
The probability of observing large
values of AE decreases very rapidly
as Bz changes to positive values
The median value of AE for Bz > 0
is less than 100, the background
level of AE from Sq variations in
measurements
CUMULATIVE PROBABILITY DISTRIBUTION FOR AE VERSUS GSM Bz
2000
1800
1600
1400
-4
1200
AE (nT)
•
1000
-3
800
-2
600
400
Median
-1
200
0
-20
-15
-10
-5
0
GSM Bz (nT)
5
10
15
20
WHAT ARE THE PHASES OF A
SUBSTORM?
• Start of Substorm - Southward turning of the IMF at dayside
magnetopause
• Growth Phase - Enhanced solar wind coupling and energy
storage in tail with ionospheric manifestations of enhanced
convection
• Pseudo Breakup - Release of small amount of stored energy
• Substorm Onset - Beginning of expansion phase
• Expansion Phase -Release of stored energy into radiation belts
(ring current), particle precipitation (aurora), Joule heating (bay
activity), plasmoids (magnetic bubble)
• Intensification - Renewal of substorm expansion that increases
disturbed region and strength of disturbances
• Recovery Phase - Reestablishment of quiet conditions and
regular structure after IMF has turned northward
• End of Substorm - Electrojet currents fade into background
SUBSTORM PHASES
A Keogram Showing Growth and Expansion
•
•
•
•
•
•
•
This figure shows a keogram plotted from a
meridian scanning photometer on the
meridian of auroral breakup.
The four Pi 2 onsets are shown by the
vertical dashed lines.
From the beginning of the data until 0408 Ut
the region of auroral luminosity is moving
southward consistent with an expansion of
the polar cap related to the storage of flux in
the tail lobes.
During this expansion there were several
intensifications of the aurora visible on the
Churchill meridian.
The 0321, 0339 and 0352 UT events
suggest pseudo breakups that did not
penetrate the poleward boundary of the
auroral oval.
The local expansion onset agrees with the
Image auroral images; the main expansion
follows the 0408 UT Pi 2 burst.
Note also that the breakup arc is not at the
equatorward edge of the aurora, and
appears to be close to the poleward edge.
•
The main onset and the expansion phase
begin at 04:06.26 UT
Cause of the DP-1 Magnetic Pertubations
•
•
•
•
•
At expansion onset a portion of the
tail current is diverted along field
lines to the ionosphere
The current is down post midnight,
westward across the auroral bulge,
and upward out of the westward
traveling surge
The ionospheric segment causes
the sudden increase in negative X
perturbations
The field-aligned segments
produce effects at midlatitudes as
indicated by diagram at bottom
This 3-D system is called the
substorm current wedge
Lobe Pressure, Convective Flow, Bz
•
Pressure (nP)
0.5
01:11:31
04:07:11
0.25
0
500
Vel (km/s)
•
A lobe pressure
enhancement
between 02-04
UT
The onset of
pressure
decrease was
associated with
strong,
convective
tailward flow and
southward Bz
The tailward
flow precedes
the major
expansion onset
at 0407 UT
0
-500
10
Bz(nT)
•
Lobe Pressure, Flow Velocity, Plasma Sheet Bz
5
0
-5
00:00
01:00
02:00
03:00
04:00
Universal Time
05:00
06:00
07:00
Dipolarization at Goes-8
•
•
•
Inclination (deg)
•
GOES-8 Magnetometer and Florida Institute of Technology Pi 2 on August 27, 2001
80
60
40
20
3
03:52:02
03:37:26
Pi 2 (nT)
•
The inclination and
Z component of B
field
Pi 2 power in
Florida
Dashed vertical
lines show Pi 2
onsets
Two minor Pi2
onsets during
growth phase
Major dipolarization
after 04:08:40 UT
Simultaneous
signatures within ½
Pi 2 cycle
04:08:40
04:17:35
2
1
0
120
100
Hp (nT)
•
80
60
40
03:30
03:40
03:50
04:00
Universal Time
04:10
04:20
04:30
WHAT CAUSES A SUBSTORM?
•
•
•
•
•
•
•
A substorm is caused by a southward turning of the interplanetary
magnetic field which initiates reconnection at the dayside
magnetopause
Dipole field lines connect with interplanetary magnetic field lines
allowing a fraction of the interplanetary electric field to penetrate the
magnetosphere and energy to be extracted from the solar wind
This coupling drives an internal convection system and associated
electrical currents which link the outer magnetosphere to the
ionosphere.
Time lags in flux and plasma transport lead to distortion of the
magnetotail and the creation of a thin current sheet close to the Earth
that becomes progressively more unstable
The unstable current sheet becomes susceptible to triggered release of
energy
An instability begins releasing energy and allowing flux and plasma to
return to the dayside
Energy in converted to flux ropes, injection into the ring current and
outer radiation belts, Joule heating of the ionosphere by steady
currents, waves, and particle precipitation
THE NEAR-EARTH NEUTRAL LINE
MODEL FOR PLASMA AND FIELD
SUBSTORM GROWTH PHASE
Dayside
Reconnection
Transpolar
Flux Transport
Erosion and
Flairing
Magnetopause
IMF
Magnetopause
New
Old
Dipole
Magnetopause Erosion
To
Sun
More
Flux
More
Flair
To
Earth
Open
Field
Line
Creation of
thin current sheet
Thinning of
plasma sheet
Plasma Sheet
Thinning
Sunward
Flow
Eart h ward
Mo t io n o f
Tail Cus p
Fo rm at io n
o f Th in
Curre nt She e t
Tail
Curre nt
FORMATION OF NENL
EXPANSION ONSET
•
•
•
•
Substorm onset begins with the
formation of a localized pair of Xand O-lines on closed field lines at
center of near-Earth plasma sheet
Reconnection creates a flux rope
centered around the O-line. The
flux rope moves tailward due to
momentum provided by forces in
X-line
Earthward flow is decelerated as it
runs into rigid dipole field.
Magnetic flux piles up creating a
turbulent region that maps to the
auroral breakup region.
The substorm current wedge is
created by the inertial current of
the flow deceleration.
Lobe Field
X-line
Plasmoid
To Distant
X-line
Auroral Brightening
and
Current Wedge
Pile up of
Earthward
Flow
Tailward
Motion
of O-line
FORMATION OF NENL
EARLY EXPANSION PHASE
•
•
•
•
•
•
Continued reconnection adds to
flux rope moving its center
further tailward.
Reconnection also adds flux to
the Earthward pile up region
creating a compression front
moving tailward.
The front maps to the poleward
edge of the expanding auroral
bulge
Closed field lines are present
between poleward edge of
aurora and X-line
Closed field lines are also
present poleward of projected Xline
The last closed field lines moves
equatorward at the speed of the
convective flow (no reconnection
of open field lines)
Lobe Field
X-line
Plasmoid
To Distant
X-line
Auroral Brightening
and
Current Wedge
Pile up of
Earthward
Flow
Tailward
Motion of
Compression
Front
Tailward
Motion
of O-line
FORMATION OF NENL
LATE EXPANSION PHASE
•
•
•
•
•
•
•
•
Reconnection moves into open
field lines of lobe
The x-line moves tailward in
discrete jumps
Earthward flows are slowed
and diverted around the Earth
The plasma sheet expands
close to Earth
A thin current sheet forms
tailward of X-line
Open field lines help propel a
macroscopic plasmoid downtail
Note compression front may
still be on previously closed
field lines
Both old and new closed field
lines are still poleward of
poleward edge of auroral bulge
Expansion of
Plasma Sheet
Plasmoid
Lobe Field
Auroral Brightening
and
Current Wedge
Pile up of
Earthward
Flow
Tailward
Motion of
Compression
Front
Tailward
Motion
of X-line
Tailward
Motion
of O-line
A SCHEMATIC SUMMARY OF THE
Pi 2 GENERATION MECHANISM
(Shiokawa et al., 1998)
CURRENT DIVERSION INTEGRATED
OVER Z
(Birn et al., Preprint, 1999)
DO SUBSTORMS HAVE MULTIPLE ONSETS OR
ARE THEIR MULTIPLE SUBSTORMS?
• Most disturbed intervals contain multiple occurrences of the
phenomena characteristic of onset
• This substructure is quasi-periodic with a period of 15-20
minutes
• Some researchers call each such event a “substorm” although
they do not include all the characteristics originally used to
define a substorm
• The question is whether these events are separate and
unrelated, or they are organized by a larger phenomenon (the
original substorm)
• If they are part of something larger, which member of the
sequence is most important and can be used as a time
reference (main onset)?
Plasma Flows Near Substorm Onset
•
Cluster Plasma Velocity in GSM Coordinates on August 27, 2001
Den (/cc)
0.4
0.3
0.2
0.1
GSM-Vx
500
0
-500
500
GSM-Vy
•
We have shown
above that the
major expansion
occurred about
04:07:34 UT
These Cluster
plasma data
indicate that a
tailward flow
characteristic of a
NENL began six
minutes earlier
It appears that this
auroral expansion
was caused by
formation of a
NENL
0
-500
500
04:07:34
GSM-Vz
•
0
-500
03:56
03:58
04:00
04:02
04:04
04:06
Universal Time
04:08
04:10
04:12
Relation of Field, Flow and Current
•
•
•
•
•
•
•
Four intervals defined
in previous figure are
shown
Blue vectors are the
field, red are the
current, black are the
flow
The anti parallel current
and field are tilted at
45 to X axis
The flow is first
orthogonal to the
direction of current and
flow
It then becomes
steadily tailward
It returns to the initial
configuration
Finally, the flow rotates
to Earthward as the
field returns to normal
orientation
Pi 2 Power on US East Coast
•
•
STACK PLOT FOR UCLA Pi 2 AMPLITUDE 08/27/01
(nT)
3
2
1
MSH Bx & By
0
4
(nT)
3
2
CLK Bx & By
1
0
2
(nT)
•
The East Coast is
close to midnight
and near statistical
center of substorms
A sequence of
progressively
stronger Pi 2 bursts
was observed
The strongest at
04:07:58
corresponds to the
poleward expansion
in aurora
What is the change
at about 0400?
1.5
1
JAX Bx & By
0.5
0
03:21:10
2
(nT)
•
03:38:32
03:52:39
04:07:58
1.5
1
FIT Bx & By
0.5
0
03:00
04:00
Universal Time
05:00
Pseudo Breakups on August 27, 2001
•
•
Two auroral intensifications accompanied by Pi 2 and localized bays were
seen at 03:39 and 03:53 prior to the main onset
These last only 1-2 images and cause no major changes in auroral oval or the
tail magnetic field
Pseudo Breakup or Main Onset?
• Auroral expansion onsets are04:00:07
often preceded by pseudo breakups
• Is the expansion onset the lingering effect of the pseudo breakup
or a new event?
• If a tail event occurs between the pseudo breakup and the main
onset which is cause and which is effect?
Saint John's Magnetogram for August 27, 2001
STACK PLOT FOR CANOPUSEW Pi 2 AMPLITUDE 08/27/01
20
03:37:11
03:51:57
2
04:07:34
1.5
-20
0400
(nT)
Bx(nT)
03:21:33
0
0400
1 stj Bx & By
-40
20
By(nT)
0.5
10
Pseudo breakup
0
5
0
4
(nT)
Bz(nT)
-10
20
0
ott Bx & By
2
-20
-40
03:00
3
1
03:15
03:30
03:45
04:00
04:15
Universal Time
04:30
04:45
05:00
03:20:58
03:38:19
03:59:33
03:52:49
04:07:31
0
03:00 03:10 03:20 03:30 03:40 03:50 04:00 04:10 04:20 04:30 04:40 04:50 05:00
Universal Time
Canada East-West Magnetometer Chain
August 27, 2001
1100
1000
Y Disturbance (nT)
900
03:39:12
03:52:55
04:06:5104:16:22
stj
Is this the
ott onset?
800
pbq
700
gil
600
rab
500
smi
400
sim
300
daw
200
mea
100
vic
Or is this it?
0
03:00 03:10 03:20 03:30 03:40 03:50 04:00 04:10 04:20 04:30 04:40 04:50 05:00
Universal Time
Discussion
•
•
•
•
•
•
•
•
Reconnection began Earthward of 20 Re six minutes before a major
substorm expansion
The x-line quickly moved outward beyond 20 Re
Earthward flow probably persisted from Cluster onset at 0401 until at least
0433
Image aurora, Gillam keogram, Florida Pi 2 – all show a major expansion
onset at 04:08 UT
Dipolarization at premidnight Goes-8 was delayed from auroral onset by ~1
minute
Dipolarization at post midnight Polar was delayed by ~12 minutes
The main onset was preceded by three Pi 2 and accompanying pseudo
breakups
There were additional short-lived auroral intensifications not evident in
distinct Pi 2
Questions for You!
Are the pseudo breakups and weak auroral
intensifications the cause of this substorm?
or
Is the major reconfiguration of the tail, high speed flows,
and loss of energy from the tail lobe the cause?
or
Did the weak auroral events cause the reconnection event?
The End!