Transcript The Magnetosphere and Plasmasphere

The Magnetosphere and
Plasmasphere
CSI 769
3rd section, Oct. – Nov. 2005
J. Guillory
•
•
Lecture 7 (Oct.18)
Bow shock scattering; Magnetosphere structure, charged particle orbits
– Advance reading: Gombosi Ch.1 & 6 (and/or Parks Ch. 4, 8 &10), Tascione
Ch.4, and Sci. Am. Apr. 1991: Collisionless Shock Waves
•
•
Lecture 8 (Oct. 25)
Dawn-dusk electric field; gyrokinetic codes; magnetotail & current sheet
– Advance reading: Gombosi Ch. 7 & 14, Parks Sec. 7.8 & 11.5), Tasc. Ch 5
•
•
Lecture 9 (Nov. 1)
MHD codes & boundary conditions; parallel E-fields & precipitation; satellite
diagnostics
– Advance reading: Gombosi Ch. 4 , Parks Sec. 7.7
•
In conjunction with the last topic:
– Joel Fedder (NRL) is scheduled for a Space Sciences Seminar on his MHD
magnetosphere code, Wed. 10/26, 3:00 p.m., 206
– Please attend.
Bow shock
Perpendicular shock
Oblique shock
waves & particles in upstream
foreshock
From Sagdeev & Kennel, Sci. Am. Apr. 1991
Foreshock region
Current in shock layer
From G. K. Parks, Physics of Space
Plasmas, AW 1991
Components of v along B and the shock surface,
for incident & reflected particles
Repeated reflections from solitons near nonsteady
shock
From Sagdeev & Kennel, Sci. Am. Apr. 1991
ISEE In-situ B-field measurements
across bow shock
Collisionless shock structure
Hybrid code simulation by M. Leroy et al, GRL 8, 1269 (1981)
C. S. Wu, D. Winske et al., Sp.Sci. Revws 37, 63 (1984)
Ion
distributions
near shock
Phase space in normal
direction
Electric potential
structure
Incoming particle scattering
• Stochastic Injection of Energetic Particles from Bow
Shock and from Tailward reconnection region
• Nonadiabatic because gyroradius ~ B scale-length
locally
• Particle orbit diffusion due to field fluctuations
• Some particles accelerated near bow shock and
magnetic reconnection regions
Magnetic field geometry
Model field topology for northward IMF
IMF +
dipole
Model field topology for southward IMF
From G. K. Parks, Physics of Space Plasmas, AW 1991
Field line motion with southward IMF (after
J. Dungey, 1961). North is DOWN in this fig.
From G. K. Parks,
Physics of Space
Plasmas, AW 1991
Detail of day-side reconnection field & flows
Field near reconnection region
(more on reconnection later)
Inner magnetosphere:
Energetic proton density contours, showing South Atlantic
anomaly
Mag. dipole offset from rotation axis
and tilted
Van Allen Belts
• Inner belt energetic protons
from cosmic ray albedo neutron decay
(CRAND) and from diffusion from
elsewhere.
• Outer belt somewhat energetic ions
from solar wind injection and
accelerated from ionosphere,
and diffusion from elsewhere.
Approx. avg. contours of spatial distribution of
trapped energetic protons & electrons
(Van Allen, 1968)
NSSDC quiet-time static empirical model
(AP-8) of energetic proton flux density
J. D. Gaffey & D. Belitza, J. Spacecraft & Rockets 31, 172 (1994)
NSSDC quiet-time static empirical model
(AE-8) of energetic electron flux density
J. D. Gaffey & D. Belitza, J. Spacecraft & Rockets 31, 172 (1994)
Calculated energetic proton lifetimes (x ne) in inner
Van Allen belt under quiescent conditions
R. C. Wentworth, “Pitch Angle Diffusion.. ”, Phys. Fluids 6, 431 (1963)
Satellite measurement of proton density vs L,
during quiet and CME-arrival conditions
OGO-5 measuremts
R. Chappell et al,
JGR 75, 50 (1970)
Squeezing the magnetosphere:
quasistatic pressure balance estimate
• Ram pressure of CME arrival + IMF
rv2
vs
• interior particle+field pressure
•Increasing B produces inductive E fields
and currents
Energy transfer from CME to
magnetosphere: time delay
From Tascione,
after Baker et
al, JGR 90,
1205 (1985)
Particle currents in the
magnetosphere
Ring current reduces B at surface
• After CME compression of day-side
magnetosphere, Bhoriz at RE decreases
~1 hr after sudden-commencement rise,
and stays reduced for 1-3 days, gradually
returning to pre-storm values.
This is correlated with injection of 10100keV magnetotail particles into ringcurrent region.
Global magnetic change index (Tascione sec 4.7)
• K: integer, 0 – 9 : 3-hr average of DB, on
quasi-log scale, for each of several ground
locations
• Kp: average of K’s from 12 locations
between 48 & 63 degrees latitude,
averaged with local & seasonal variation
filtered out
Dst index (Tascione p.51)
• Hourly avg. (from 4 low-latitude ground stations)
of changes in horizontal component of B, with
seasonal variations subtracted out.
• A measure of changes in ring-current intensity
AE (Auroral electrojet) index
• Spread between max positive & max negative
changes in horizontal component of B at several
auroral-latitude locations, 62.5 – 71.6 degree
latitudes
• Global AE = maximum of the positive changes
in horizontal component at any such location maximum of the negative changes in horizontal
component at any such location
Electric fields and more currents &
plasma flows
• Plasma corotation-induced E field:
–
–
E = - (w x r) x B
= BoRE (RE/r)2 (2sin l el + cos l er)
approximately, for dipole B.
• Inner part of earth’s magnetosphere
corotates.
•Added to dawn-dusk E field due to solar wind
Model E-fields in equatorial plane
Collisionless plasma flow (not
currents) in E perpendicular to B
• vD/c = E x B /B2
(cgs) , if (as usual) E (cgs) << B (G)
B-Field-aligned currents and fields
• High-conductivity acceleration-limited
currents along B-lines
• Downstreaming charges arrive & die at
dense ionosphere, producing auroral glow.
• Upstreaming charges from ionosphere
populate plasmasphere and mag’sphere.
From Tascione
Field-aligned ion beam distributions in plasma sheet
boundary layer from ISEE-1, 16 Feb 1980.
(T.E. Eastman, R.J. DeCoster & L.A. Frank, in Cross-Scale
Coupling in Space Plasmas
Velocities for steady-state polar wind with no
field-aligned current
S. B. Ganguli, H. B. Mitchell, &
P. Palmadesso, NRL Memo
Report 5673, 1985
Velocities (magnitude) 70 min after onset of
a current of -1 mA/m2 at 1500 km
S. B. Ganguli, H.
B. Mitchell, & P.
Palmadesso, NRL
Memo Report
5673, 1985
Heating by these currents
S. B. Ganguli,
H. B. Mitchell, &
P. Palmadesso,
NRL Memo
Report 5673,
1985
Particle orbits in inner
magnetosphere
• Assumptions:
DB/B is small over a gyroradius &
gyroperiod
Motion of the particles of interest is
collisionless (except if the hit the
ionosphere, where they die)
Charged particle orbits, static B:
(for B quasistatic and gradB/ B << rg-1 )
• Fast gyration about field line
• North-south bounce due to “magnetic mirror”
force
• Slow east-west drift due to inhomogeneous
magnetic field
• ExB drifts due to electric fields
Gyration about B-line:
• rg = gmv/qB (MKS units)
Wc = qB/gm (MKS units) or qB/mc (cgs)
Sub-kHz to MHz angular frequencies (2pf):
Wc ~ 1.76x107 B(G)/g for electrons
~ 104 B(G) for protons
Scale of Proton Gyroradius &
Gyrofrequency
•
0.1 G
•
.01 G
rg(1 MeV) = 10 km A (eperp/e)½
wci = 1000 rad/s; f ci = 160/s
rg(1 MeV) = 100 km A (eperp/e)½
wci = 100 rad/s; f ci = 16/s
tN-S ~ 1.3 s (L=2) ~50 gyroperiods
Adiabatic invariants
•
General form
∫pdq
• 1. Magnetic moment invariant
gm = pperp2/2mB
(m = qrg2/2c)
• 2. Bounce invariant J = ∫ppards
• 3. Longitudinal drift invariant (L shell)
North-South bounce motion
• Determined by pitch-angle of fast velocity
vector at magnetic equator,
• And by energy conservation.
• If Eparallel =0, each particle’s parallel energy
is converted to perpendicular energy until
it has no more parallel momentum, then it
reverses its parallel motion.
Effective magnetic mirror force
• When a very-small-size dipole moves along
magnetic field lines of an inhomogeneous field,
there is an effective “force” parallel to the field
line that has magnitude & sign
Fparallel = - m • d B/ds
where s measures arclength along the magnetic
field. A dipole entering a region of stronger
magnetic field thus has a retarding force on it,
slowing its parallel motion.
• One may ask “How can this be? The magnetic force on
a charged particle, qvxB/c, is always perpendicular to v
and so can do no work on the particle if B is constant in
time.”
•
In fact, the particle kinetic energy,
½ m vz 2 + ½ m vy 2 , does not change; only the partition
between vz and vy changes.
• And this change of the direction of v is due to the fact
that the particle is not exactly at the position of its
guiding center, so the directions of the field lines of B at
the particle are not quite the same on opposite sides of
the gyro-orbit, leading to a gyro-averaged vxB force that
has a parallel component.
• The constancy of m = (½ m vperp 2 ) /B during collisionless
nonrelativistic charged particle motion along B, and
the constancy of ½ m vpar 2 + ½ m vperp 2 = KE,
mean that vpar can be expressed in terms of its value at
some reference point so by
•
½ m vpar 2 = (½ m vpar 2)o - m(B - Bo),
• i. e. the parallel motion is derivable from a potential, mB.
• (If there is also a static electric field parallel to the
magnetic field, the effective potential for parallel motion
of the “dipoles” generalizes nicely to mB + qf .)
• When B has a minimum at some reference point so
along each magnetic field line encircled (or “enhelixed”)
by a particle, the collisionless parallel motion will be that
of a particle in an effective potential well (remember,
though, that the magnetic potential depends on the
constant m, which is not the same for all particles !).
Half of Loss-cone(s) in magneticequator velocity-space
Shown for no parallel electric field
Particle turns around where ( & if) vpar 2 = 0,
where all the energy is converted to
perpendicular energy,
i. e. at B such that
(½ m vpar 2)o - m(B - Bo) - q(f - f o) = 0.
This turning point equation, with the help of
magnetic moment constancy
(½ m vperp 2 ) /B = (½ m vperp 2 ) o /B o ,
specifies the turning points as where
(½ m vpar 2)o - (½ m vperp 2 ) o (B/Bo - 1) - q(f - f o) = 0.
• When there is negligible parallel electric field this is
simply
B /Bo = 1 + (vpar2 / vperp 2)o ,
• so each trapped collisionless particle mirrors, i.e.
changes its sign of parallel velocity, at a value of B/Bo
that depends on its pitch angle at the minimum of the
magnetic field.
• In the reference-plane velocity space {vpar o, vperp o}, one
can draw a boundary for any value of B1, such that
particles with |v perp o | above the boundary will be
trapped in the spatial region where B < B1 , and those
with |v perp o | below the boundary can progress to higher
values of B than B1 if nothing else stops them first.
Loss regions in midplane velocity space(s) when
there is a steady parallel E field toward the
ionosphere (positive potential on field line)
trapped
Lost to ionosphere
in 1 bounce
Positive potential occurs so as to reduce electron
loss rate to the (increased) ion loss rate.
Same thing in midplane energy space
Particle gyration & bounce in inner
magnetosphere
From T. Tascione, Intro to the Space Environment
(From G. K. Parks,
Physics of Space
Plasmas)
Energy conservation (with E=0)
ao = pitch angle at mag. Equator (l = 0)
Bo = field strength at mag. Equator
Charged particle longitudinal drift due to
magnetic field inhomogeneity
Cross-field drift of – and + particles under force F
From Parks, Physics of Space Plasmas
Longitude-drift periods (from Parks)
Drift Rate (in terms of energy, mag. Moment,
bounce invariant, bounce period, and L)
• For static magnetic dipole, with E = 0:
•
<df/dt> = - (2cLRE e /em) (3/2 - J/4et)
•
with t = ∫ds/vy = N-S bounce period,
•
and J = m ∫vy ds = bounce action integral.
• T. G. Northrop, in Radiation Trapped in the Earth’s Magnetic Field,
B. M. McCormac, ed., Reidel 1966
• For nonstatic dipole B without shear
(but changes slow enough to preserve J and m):
• <df/dt> = - (2c/et)∫rdq (B/Bq ){(2m(H - mB - qF))½
[(Bq /B)( ∂/∂L)(rB/Bq) + ½ r ∂B/∂L ]
- ½ (2m(H - mB - qF))-½ [2m(H -qF)r ∂lnB/∂L+ rq ∂F/∂L]}
with F = electric potential
and f = longitude angle.
• T. J. Birmingham, “Guiding center drifts in timedependent meridional magnetic fields”,
Phys. Fluids 11, 2749 (1968)
Ring Current from drifting, gyrating particles
(Parks sec. 7.7.4, with corrections)
• (a) From guiding-center drifts
Jgc = e (ni vdrift(i) - ne vdrift(e-) )
= Σn(KE) e [(KE/e) (1 + cos2a) bxgradB /B2]
bxgradB/ B2 = -3/rB if at l = 0
• so
•
Jgc ~ -3n[<KE>i + <KE>e ]/rB
(for equatorial particles)
Igc = ∫JgcdV /2pr ~ - 3Etot/(2pr2B)
• (b) From pressure gradient of gyrating particles
•
JgradPperp xB = grad (n[<KE>i + <KE>e ]perp.)
•
IgradPperp ~∫rdrdl grad (n[<KE>i + <KE>e ]perp.)/B
•
~ + Etot/2pr2B
• so this (b) current reduces the average net ring current magnetic
field by about 1/3. The net ring current then reduces the magnetic
field at the magnetic equator at 1 RE by
•
DB/B = - (2/3) Etot/Emag ,
where Emag is the volume-integrated energy in magnetic field.
• See more general derivation in R. L. Carovillano & J. J. Maguire, in
Physics of the Magnetosphere, (Carovillano et al, ed’s), Reidel,
1968.
• Ring current usually peaks at 4-5 RE (quiet); at 2-4 RE (storm)
• Mean proton energy: 85keV (90% are in 10 - 250 keV)
• Quiet-time ring current density ~ 10-8 A/m,
increased by factor of several during storms.
• See Tascione sections 5.4.3, 5.9, 5.10
Typical energy spectrum of
energetic protons
Power delivered by solar wind/ CME
• Power = Current x ∫(-vxB)•dl
• Current varies as wc, i.e. as B
•
Power varies as B2v sin4(q/2)
where
 q = angle of IMF from northward
sin = 0 for northward IMF
sin = 1 for southward IMF
• J. K. Alexander, L.F. Bargatze, J. L. Burch et al.,
“Coupling of the solar wind to the magnetosphere”
in
Solar Terrestrial Physics
D.M. Butler & K. Papadopoulos, ed’s. NASA, 1984
• Tascione, sections 3.7, 5.8, 5.10
Energy injection into ring current
• Empirical approximate formula for ring-current
addition rate in terms of Dst and ring-currentenhancement lifetime t:
UR(J/hr) = 4x1010(dDst/dt + Dst/ t)
(Tascione sec.5.10)
See Akasofu [Sp. Sci Rev, 28, p160, 1981] for a related
formula:
|Dst| ~ 60*(log [epsilon] - 18)**2 + 25
where epsilon = B2v sin4(q/2)
Nov. 6, 2001 event
• Southward B component ~80 nT
• Unusually sharp CME shock with speed >1000km/s
• Nearly perpendicular shock
• L=8 SEP’s showed sharp rise in # on shock arrival
• L=3: 14-25 MeV protons arrived minutes before shock
and were trapped when shock arrived
via front-side & cusp entry
stayed trapped til Oct ‘03 storm detrapped them
• 3-20 MeV electrons enhanced at first, but
deep dropout of total >1MeV electron flux at L=3-8,
with few-days recovery time
• Mary Hudson’s PIC particle follower, riding on
Fedder-Lyons-Mobarry MHD code,
followed particles from ACE input data
• Cluster data (Morikis & Kistler, UNH)
Cluster apogee 20 RE, perigee 4 RE, every ~48 hrs
50 hr orbit, 2hrs in magnetosphere at ~4RE
• Sampex data: ~1-3 MeV electrons, 10-20 MeV
electrons
Stochastic Injection of Energetic Particles
from Bow Shock and
Tailward reconnection region
• Nonadiabatic because gyroradius ~ B scale-length locally
• Timescale t varies as m5/4e-1/2
• Flux density injected varies as density at low densities
• M. G. Rusbridge, “Non-adiabatic effects in charged-particle
motion near a neutral line”, Plasma Physics 19, 1087 (1977)
and
• “Non-adiabatic charged particle motion near a magnetic field
zero line”, Plasma Physics 13, 977 (1971)
• W. Peter & N. Rostoker, “Theory of plasma injection into a
magnetic field”, Phys. Fluids 25, 730 (1982)
• J. Chen & P. J. Palmadesso, “Chaos and nonlinear
dynamics of single-particle orbits in a magnetotail-like
magnetic field”, JGR 91, 1499 (1986); errata 91, 9025 (1986)
Particle Diffusion
Dominated by field fluctuations in storm conditions.
Lee/Sydora Gyrokinetic Code calculates for Tokamaks.
Diffusion model:
W. N. Spjeldvik, “Consequences of the duration of solar
energetic particle-associated magnetic storms on the intensity of
geomagnetically trapped protons”, in Modeling Magnetospheric
Plasma, T.E. Moore & J.H. Waite, ed’s. AGU 1988
J.M. Cornwall, “ Radial diffusion of ionized helium and protons: a
probe for magnetospheric dynamics” JGR 77, 1756 (1972)
df/dt = L2d/dL (DLLL-2df/dL) - Af + Gm-1/2df/dm
A = charge exchange factor, G = Coulomb slowing
DLL(L, m) given in Cornwall (1972),
assumes power-law ( n-2) spectrum of fluctuations in B and E.
Flow dynamics of charge-neutralized plasma
fluid:
• [∂t + U•grad]U = (1/rmo)[(B•grad)B - grad(B2/2)] - (1/r) divP + g
P = pressure tensor = pperp I + (ppar - pperp)bb
(div P )perp = gradperp pperp - (ppar - pperp)(b•grad)b
(div P )par = (b•grad)ppar + (ppar - pperp)divb
div P = gradp for isotropic pressure
• [∂t + U•grad] (pperp/rB) = 0
• [∂t + U•grad]( pparB2/r3) = 0
• G.F. Chew, M.L. Goldberger, & F.E. Low, Proc. Roy. Soc (Lon.)
A236, 112 (1956)
• N.A. Krall & A.W. Trivelpiece, Principles of Plasma Physics, McGraw
Hill 1973
Magnetosphere Simulation
• Particle codes, incl. gyro-averaged particle followers (e.g.
Mary Hudson’s at NASA & R. M. Winglee code at UW)
• Fluid (MHD) codes
–
–
–
–
Fedder-Lyon-Mobarry code (NRL)
BATSRUS (U. Michigan)
Spicer code(s): Odin etc.
Modified MHD: Winglee
• Hybrid (particles and MHD) codes
– Rice MSM code
– Kazeminezhad 2D code
• Models are available for community use:
– CCMC: http://ccmc.gsfc.nasa.gov/
– UCLA: http://www-ggcm2.igpp.ucla.edu/
• Source codes in public domain:
– GEDAS (Japan, T. Ogino) (Japan, T.)
http://gedas22.stelab.nagoyau.ac.jp/simulation/jst2k/hpf02.html
– BATSRUS: http://csem.engin.umich.edu/
– NRL: http://www.lcp.nrl.navy.mil/hpcc-ess/software.html
• FCTMHD3D (C.R. DeVore)
• AMRMHD3D (P. MacNeice)
– Zeus 3D MHD (Michael Norman):
http://zeus.ncsa.uiuc.edu:8080/lca_intro_zeus3d.html
– CFD Codes: http://icemcfd.com/cfd/CFD_codes.html
Fedder-Lyon-Mobarry (FLM) Code:
distorted spherical coord. grid
MHD eqns as solved in FLM code
J. G. Lyon, “Numerical
methods used…”, Proc.
ISSS-7, 26-31 March
2005
FLM Code
• Does not include particle acceleration (since it’s
an MHD code)
• but does show overall energetics of CME
coupling for southward IMF,
• and shows very weak coupling for northward
IMF.
• Coupling is by fast magnetosonic wave
propagation from magnetopause.
• Shows Poynting vector energy flow from these
waves.
BATS-R-US Code (U. Mich.)
Block-Adaptive-Tree Solarwind Roe-Upwind Scheme)
•
•
•
•
•
Gombosi et al
3D MHD, Eulerian xyz grid (x toward sun)
Block-adaptive mesh refinement
Cell-centered finite volume method
Upwind-differencing Riemann solver
(Powell 1994)
• Efficiently parallelized
• High computation/communication ratio
• Runs on Sun, SGI shared memory, Cray
T3D & T3E, and IBM SP2
• Simulation box typically 192 RE wide,
+192 to -384 RE in x direction
• Cell size typically .25 RE to 32 RE
• Inner boundary at 3 RE (no mass flow
across it) coupled along assumed dipole B
lines to finite tensor conductivity, heightintegrated ionosphere layer at 1 RE [M. L.
Goodman, Ann. Geophys. 13, 843 (1995)]
• Dipole inner field separated off [as in
Tanaka, JGR 100, 12057 (1995)]
BATSRUS simulation of outermost closed B
lines, for Parker spiral IMF
Winglee modified MHD code
• R.M. Winglee, “Regional Particle simulations and Global
Two-fluid Modeling of Magnetospheric Current Systems”,
in
J. L. Horowitz et al., Cross Scale Coupling in Space
Plasmas, QC 809.P5 C76, 1995
• Uses a 2-fluid modified MHD set of equations
• Gets the injection of currents & plasma across B-field
lines
Rice MSM Code
Rice MSM Code
• E. C. Roelof, B. H. Mauk, R. R. Meier, K. R. Moore, & R.
A. Wolf, “Simulation of EUV and ENA magnetospheric
images based on the Rice Convection Model”,
in Instrumentation for Magnetospheric Imagery II, SPIE
1993.
(ENA = energetic neutral atom)
• Streamlined version of RCM = MSM (magnetic
specification model), has non-self-consistent E field from
“phenomenological convection patterns”.
F. Kazeminezhad new code
• 2D hybrid
Triangular
finite-element
grid
MagnetoTail
Magnetic Reconnection
Modeling driven reconnection
2-D Compressible Resistive MHD
Simulation of
Driven Reconnection
S. -P. Jin & W. -H. Ip, Phys. Fluids B3, 1927 (Aug. 1991)
• Plasma beta at inflow boundary of simulation box: initially 0.1
• Alfven Mach # of inflow: MA = 0.15
(for -.5 < z <+.5), tapering to 0
at |z| >1
• High Lundquist Number: 400 - 2500 (very low resistivity)
– Lundquist Number = ratio of JxB force to resistive mag. diffusion force
•
•
•
•
Initial Bz(x) profile: half sine wave -w < x < w (w <1), 1 for |x| > w
(odd function of x)
Initial state in pressure balance
Grid resolution in x: Dx increases 13% every step.
Grid concentrated in center near x = 0.
Time in units of Alfven-wave x-crossing time. Sim. ~ 40 units.
Implicit integration scheme: Y. Q. Hu, J. Comp. Phys 84, 441 (1989)
B lines, v vectors, DT(%), Dr(%)
Time ↓
S-P. Jin & W-H. Ip. 2D compressible MHD sim. , PhysFluids B 3, 1927 (1991)
PIC simulation of particle orbits near a
magnetic reconnection line
• H-J. Deeg, J.E. Borovsky & N. Duric, Phys Fluids B 3,
2660 (1991)
• Geometry and results shown in following slides
Region where “magnetic insulation” fails, i.e
where B is weak
H-J. Deeg, J.E. Borovsky & N. Duric, Phys Fluids B 3, 2660 (1991)
Geometry for PIC simulation of particle
acceleration near reconnection region
H-J. Deeg, J.E. Borovsky & N. Duric, Phys Fluids B 3, 2660 (1991)
Proton orbits in views 1 & 2
Proton orbit in views 2 & 3
Energy gain of protons entering near neutral point
H-J. Deeg, J.E. Borovsky & N. Duric, Phys Fluids B 3, 2660 (1991)
Final proton energy vs initial proton energy, for
protons initially incoming near neutral point
H-J. Deeg, J.E. Borovsky & N. Duric, Phys Fluids B 3, 2660 (1991)
Turbulence in B-line reconnection
Matthaeus & Lamkin, PhysFluids 29, 2513 (1986)
Magnetic
field
Fluid
streamlines
Contours
of
constant
J
Contours
of
constant
vorticity
Disturbed magnetotail reconnection at current
sheet can launch plasmoids & relax (as well as
accelerating particles forward & backward)
E. W. Hones, Sci.
Am. March 1986
Some references on field-line reconnection
• Observations by Cluster satellite:
A. Runov et al., Geophys. Res. Lett. 30, 1579 (2003)
• Observations by WIND satellite:
T. D. Phan et al., Nature 404, 848 (2000) ;
M. Oieroset, R. P. Lin et al., Nature 412, 414 (2001)
• 3D PIC simulation:
P.L. Pritchett & F. Coroniti, JGR 109, A 01220 (2004)
• 2D simulation with “guide field” normal to plane:
P. L. Pritchett (UCLA): “Onset & Saturation of Guide-field
Magnetic Reconnection”, Phys. Plasmas 12, 062301 (June
2005)
More references on field-line reconnection
• Particle acceleration & orbits:
H-J Deeg, J.E. Borovsky & N. Duric (LANL), “Particle
acceleration near X-type magnetic neutral lines”, Phys.
Fluids B 3, 2660 (1991)
• Electric field enhancements (EFE):
J. D. Scudder & F. S. Mozer, “Electron demagnetization
and collisionless magnetic reconnection in b<<1
plasmas”, Phys. Plasmas 12, 092903 ( Sept. 2005)
• Role of microinstabilities (anomalous resistivity):
M. Ugai & L. Zheng, “Conditions for fast reconnection
mechanism in 3D” Phys. Plasmas 12, --- ( Sept. 2005)
Satellite sensors
•
Radiation Belt Mappers
•
GOES (ESA)
•
Cluster, Vortex
•
Doublestar
•
Polar
•
Image
•
Geotail (Japan)
•
ISEE1-3, IMP1-8 & other former sats with elderly data
•
Ionospheric satellites measuring energetic particles: DMSP, SAMPEX
etc.
•
Upcoming: NPOESS & NPP
Living With A Star Research Network
Pole Sitter
Solar Dynamics Observatory
L1 Solar Sentinel
Ionospheric
Mappers
L2
Radiation Belt
Mappers
Distributed network of spacecraft providing continuous observations
•Geospace Dynamics Nework with constellations of smallsats in key regions of geospace.
How to find satellite orbit info (& related data)
• http://pwg.gsfc.nasa.gov/orbits
/menu_orbits.html
• Orbits for Wind, ISTP, Cluster, Image, Polar
GOES description
• GOES (Geostationary Operational Environmental Satellites,
NOAA/NESDIS)
• 2 spacecraft at 75deg W and 135deg W, one at 98deg W and/to
108deg W, moved with season.
• 35,600 km equatorial orbit, spin axis parallel to earth’s spin axis.
Telemetry to NOAA ERL, Boulder.
• measuring:
•
solar X-rays,
•
B field at satellite,
• high energy particles, via SEM (Space Environment Monitor).
• SEM has
• (a) Total Energy Detector (TED)- intensity of energetic particles 0.320 keV in 11 bands;
• (b)Medium-Energy Proton & Electron Detector (MEPED) - 30 keV60MeV;
• (c) High-Energy Proton & Alpha Detector (HEPAD) - 370 MeV- >850
MeV.
Cluster & Vortex
Cluster & Doublestar (DSP)
Cluster data (Morikis & Kistler, UNH)
Cluster: (ESA & NASA, 2000)
Cluster apogee 20 RE, perigee 4 RE, every ~48 hrs
50 hr orbit, 2hrs in magnetosphere at ~4RE
Some Cluster results
• Cluster has now proven the existence of The
Kelvin-Helmholtz instability as an important solar
wind entry process.
• These large-scale vortices could lead to
substantial entry of solar wind to populate the
Earth's magnetosphere. (Tai Phan, UCB
SpSciLab.)
Polar: orbit
(http://pwg.gsfc.nasa.gov/orbits/aaareadme_polarpar.html
The POLAR orbital
parameter plots show the
radial distance, eccentric
dipole (ED) magnetic local
time (MLT), and eccentric
dipole L-shell value. The
darker segments
correspond to times when
one of the magnetic
footpoints (traced down to
100 km altitude using the
T89, Kp=3-,3,3+, model)
falls in one of the following
regions: cusp, cleft, or
auroral oval.
Polar: observation of an event
Images in visible light from the Polar satellite's Visible Imaging System
compares the northern auroral regions on May 11, 1999, and a more
typical day on November 13, 1999. Credit: University of Iowa/NASA.
• Polar, cont’d
• May 11, 1999 event: solar wind flux dropped a lot
• produced an intense "polar rain" of electrons over one of
the polar caps of Earth.
• Electrons flow unimpeded along the Sun's magnetic field
lines to Earth and precipitate directly into the polar caps,
inside the normal auroral oval.
• Such a polar rain event was observed for the first time in
May 1999 when Polar detected a steady glow over the
North Pole in X-ray images.
• Jack Scudder, U. Iowa, PI for the
Hot Plasma Analyzer on NASA's Polar spacecraft.
Scudder and Don Fairfield of Goddard had predicted the
details
• In parallel with the polar rain event, Earth's
magnetosphere swelled to five to six times its
normal size.
• NASA's Wind, IMP-8, and Lunar Prospector
spacecraft, the Russian INTERBALL satellite and the
Japanese Geotail satellite observed the most distant
bow shock ever recorded by satellites.
• SAMPEX spacecraft reveal that in the wake of this
event, Earth's outer electron radiation belts dissipated
and were severely depleted for several months
afterward.
Image Satellite ENA sensors
Image website (Southwest Research)
• http://pluto.space.swri.edu/IMAGE/
• HENA: D. G. Mitchell and HENA team, the
Johns Hopkins University Applied Physics
Laboratory
• MENA: C. J. Pollock and J.-M. Jahn, Southwest
Research Institute
HENA Images of ENA fluxes during the
July 15-16 2000 Geomagnetic Storm
Geotail (Japanese space program)
• Instruments:
Solar wind, hot plasma, & composition
analyzers,
directional data on electrons/protons/helium
above 20keV, protons above 400keV, electrons
above 120kev, B field, etc.
• http://wwwistp.gsfc.nasa.gov/istp/geotail/geotail_key_para
meters.html
DMSP Satellites:
• Orbits: circular, sun-synchronous, polar, ~850km alt.
• 98.7 deg inclination, period 101 min., revisit time 6 hrs.
•
Global coverage @ 12hrs each satellite
• Communications: S-band, about 3 MBPS in 1995;
maybe more capacity now.
• Design life: 3-5 yrs.
• Block(group) 5D-2 (5 sats) launched 1991-98, earlier
ones presumably now down or inoperative;
•
• Block 5D-3 (5 more satellites, S16-20, built by Martin
Marietta) launched 1999-06; Block 6 beginning 04.
DMSP, cont’d.
• Relevant sensors for space weather:
• SSI/ES Ionospheric Plasma Drift/Scintillation Monitor: 4 sensors
monitoring ion & electron densities, temperatures, drift velocities of
ions, and plasma irregularities above the F region.
SSI/ES-2, 3 are enhanced versions, flown since ‘94 and ‘99.
• SSJ/4 Precipitating Electron/Proton Spectrometer
• SSB/X: X-ray detector array - x-rays from earth’s atmosphere.
Upgraded version SSB/X-2 can also detect gamma ray bursts.
• SSM: magnetometer measures B-field fluctuations due to hi-latitude
ionosphere currents.
Sampex (GSFC)
• Solar Anomalous and Magnetospheric Particle Explorer
(Medium Earth Orbit).
• First of NASA's Small Explorer (SMEX) missions.
• Typical orbit: 520 x 670 km, 82 deg inclination
• Energy, composition and charge states of :
(1) cosmic rays
(2) solar energetic particles
(3) magnetospheric electrons trapped by the Earth's
magnetic field).
• http://www.astronautix.com/craft/sampex.htm etc.
• Sampex data: ~1-3 MeV electrons, 10-20 MeV electrons
• PET: Proton-Electron Telescope: energy spectra of
electrons from 0.5 to 30 MeV, and of H and He from ~ 20
to 200 MeV/nuc
• http://www.srl.caltech.edu/sampex/
Upcoming: NPP & NPOESS
• The NPP satellite is scheduled for launch in 2007 into a circular
sun-synchronous polar orbit at a nominal altitude of 824 kilometers
and a 10:30 a.m. descending node.
• This orbit provides a 16-day repeat cycle (8-day quasi-repeat),
similar to that of the EOS satellites.
• Ref.:The NPOESS Preparatory Project: Architecture and
Prototype Studies (Aerospace Corp. website)
• The National Polar-orbiting Operational Environmental Satellite
System (NPOESS) represents a convergence of systems previously
operated by the Department of Defense and the National Oceanic
and Atmospheric Administration (NOAA).
• Scheduled for launch in 2009, it will support a broad range of
activities in global environmental monitoring, meteorology, and
climatology.
NASA CDAW at GMU, Mar. 2005
• http://cdaw.gsfc.nasa.gov/geomag_cdaw
/register/wg2_participants.html
Names & contact information of researchers in
magnetosphere dynamics & data
•http://solar.scs.gmu.edu/meetings/cdaw/data/
cdaw2/wg2_datatable.htm
Data files for selected events, from several satellite
instruments (click on “data” & first “WG2 data table”)
Magnetosphere Homework
Assignment, 10/25/05
•
1. Look up typical magnetotail storm-period data (Bfield strength, particle densities, particle
“temperatures”) from, e.g., IMP 8 data.
•
•
•
•
•
2. Use these data along with Fig. 5.6 of Tascione to estimate the order of magnitude of:
(a) tailward speed of ejected plasmoid (km/s)
(b) directed particle energy of tailward-ejected plasmoid (J)
(c) kinetic power loss (mean particle energy loss rate) during plasmoid ejection (W)
(d) magnetic energy stored in magnetotail (J)
•
3. Use ACE or WIND data to estimate the typical order of magnitude of CME ram pressure rv2
(J/m3) and of CME-enhanced power delivery to day-side magnetopause (W), for southward Bz = 80nT and twice the typical Parker-spiral westward By. Is this pressure much bigger than the
magnetic field pressure? Estimate the power (W) delivered into the magnetopause by such a
CME.
•
4. Tascione problem 5-4
•
5. Tascione problem 5-5
•
6. Tascione problem 5-12
CSI 769 Class Project, fall 2005
Magnetosphere portion:
•
This part of the project focuses on the energetics of the Halloween ‘03 CME-induced changes in
the magnetosphere, by doing five short order-of-magnitude calculations based on retrieved data.
•
1. From Wind or ACE data, estimate the peak CME (particle + magnetic) pressure increase on the bow
shock, and its rise rate during the Halloween ‘03 event.
2. From earthbound magnetometer data, e. g. Dst, estimate
(a) the time delay of surface DB after the bow-shock energy delivery, and
(b) the energy and power delivery to the enhanced ring current during the storm.
(c) If the time delay is related to propagation of a disturbance at near the Alfven speed, use the magnitudes
of B and estimated plasma densities to compare the time delay to that of the most direct delivery route.
(d) Is the ratio of estimated change in ring-current energy (volume integral of DKE + D(B2/2mo)) to CME
energy (magnetopause-intersecting volume integral of energy density in the CME on bow-shock arrival) of
order unity or <<1?
3. (a) Based on your estimates of magnetospheric DB due to enhanced ring current and its risetime, estimate
the peak E fields (mV/m) induced, and compare them to the corotation E field.
(b) Give an estimate of the peak E field on the topside of the ionosphere, say at 500km altitude, and the
ExB drift speed E/B (km/s) at 60degrees magnetic latitude.
4. Use geotail data during the storm to estimate the peak change in magnetic energy storage in the
magnetotail volume, and its buildup rate. Compare these numbers with the estimated frontside energy
arrival by the CME.
5. Use NOAA energetic-particle flux data etc. to estimate the change in total energy in MeV (and higherenergy) protons transported by the storm to the auroral ionosphere, and compare this with the other energies
calculated above.
Part of the data is collected at “http://solar.scs.gmu.edu/meetings/cdaw/Data_master_table.html”
•
•
•
•
•
•
•
Some textbook errata
• Tascione
• Eq. 1.17:
• Eq. 1.33:
•
•
•
•
•
•
= (not +) after first term
B(vector)x gradient of scalar B (magnitude of vector B),
not of vector B.
Fig. 2.6:
protons don’t arrive with predominantly 45 degree
Incidence, even though B does. Water-sprinkler effect.
p. 35:
U components: theta & phi here are interchanged from
the usual (i.e. Jackson).
p.38 Eq 3.28:
factor of d is ignored in the final proportionality
and is treated as constant in 3.29, but
reappears as Lo2 in 3.30.
p.44 Eq. 4.9:
Z on left, not H.
Eq. 4.10:
B on left, not H.
p. 59 Eq. 5.20:
+ sign (not -) in numerator.
Some textbook errata, cont’d.
•
Parks
•
•
•
•
p. 56 Eq. 3.36 & 3.37: Confused notation. r and lambda are component indices, not independent
variables.
p. 72 Eq. 3.73: Careful! The rotation axis is not the magnetic axis. See Tascione Eq’s 4.1 & 4.2.
p. 106, first line of sec. 4.55.6: current density, not currents. Current is meaningful for single
charged particle in motion (I=qv). Current density is not
p. 139 problem 18: dimensional error in formula.
p. 156, top two eq’ns: either one or the other (not both, unless gamma = 1).
p. 249 Eq. 7.20: see eq. 7.57 when p is anisotropic.
p.255 below eq. 7.39: “outward” = out of paper (as looking down from N pole), not radially
outward from earth.
p. 259 after eq. 7.53: del parallel plus del perp. (not -)
p.261 after eq. 7.65: B is not necessarily given, just static.
p.264 after eq. 7.70: Br vanishes at the magnetic equator (only).
p.265 eq. 7.74: sum over species! Epsilon is the energy-density of all the drifting particles (e + i).
p.267 Eq. 7.82: delta BT/Bs on left side, not delta BT.
•
•
•
•
•
•
•
•
•
•
p. 314, before sec. 8.2.2: Plasmas in steady state do support free charges, but mainly at or near
their boundaries. Like a pretty-good conductor, they move the net charge to the ‘surface”.