Tutorial on Inner Magnetospheric Electric Fields, GEM Workshop, 6/021 Inner Magnetospheric Electric...

Tutorial on Inner Magnetospheric Electric Fields, GEM Workshop, 6/02 1

Inner Magnetospheric Electric Fields

R. A. Wolf

Rice University

(Unpublished results provided by Stanislav Sazykin, Mei-Ching Fok, Trevor Garner, and

Jerry Goldstein)


Introduction and Outline

• Introduction/Outline

• Neutral Wind Effects

• What Laws Govern Magnetospheric Generators of Inner Magnetospheric E?

• Shielding of the Inner Magnetosphere–Basic Physics

• Overshielding–Theory and Observations

• Undershielding–Theory and Observations– Asymmetry of Main-Phase Ring Current

– Large Eastward Electric Fields in Post-Dusk Equatorial Ionosphere

• Subauroral Polarization Streams and SAID Events

• Summary


Inner Magnetospheric Electric Fields• Limitations:

– By “inner magnetosphere,” I mean the region earthward and equatorward of the electron plasma sheet.

• Maps to the subauroral ionosphere

– This talk emphasizes potential electric fields, not induction.• Consider time scales > a few minutes. No waves.• I assume that E= in ionosphere, for these time scales.• Also assume that is constant along field lines. (We’re talking about the subauroral

ionosphere here.)

– I’m going to offer only a very quick summary of neutral-wind effects.

• Why are inner magnetospheric potential electric fields important?– They control plasmasphere and ring-current dynamics. We won’t understand ring

current injection until we understand the associated electric fields.– They affect the radiation belts somewhat (though induction electric fields, and

waves are probably more important).– They drive much of the dynamics of the low- and mid-latitude ionosphere

• They seem to be the most unknown element in modern ionospheric modeling of the subauroral ionosphere.

• Disruptions of the mid- and low-latitude ionosphere seem to be the most important aspects of space weather at present, particularly for the military.


Quick Summary of Neutral-Wind-Driven Inner-Magnetospheric E Fields

• Fields driven by winds resulting from solar heating of the day side. These drive the Solar Quiet (Sq) currents and are pretty well understood, quantitatively.

• Dynamic processes are only partly understood:– Disturbance dynamo winds are equatorward and westward winds driven by heating

of the auroral zone in storms. They propagate to equatorial region in ~ 1 day. There still isn’t convincing agreement between the best models and observations.

– Tidal winds propagate up from lower atmosphere.– Traveling ionospheric disturbances are gravity waves that propagate equatorward,

driven by auroral zone heating.

• Neutral-wind effects probably dominate inner-magnetospheric E fields in times of geomagnetic quiet and they are significant in active times.

• See tutorial reviews by Richmond (in Solar-Terrestrial Physics, ed. Carovillano and Forbes, Reidel, 1983) and (in Handbook of Atmospheric Electrodynamics, Vol. II., Ed. H. Volland, CRC Press, 1995)


Magnetospherically Driven Inner-Magnetospheric Electric Fields

• The large-scale electric field in the magnetosphere is that of magnetospheric convection.– Simplest approximation:

€

=−Eoy−BoωERE

3

rUniform dawn-dusk field Corotation

– Next simplest approximation–Stern-Volland and corotation

€

=−Ayr−BoωERE3

r• Neither of these simple models captures the complexity of

the real system. This talk will concern effects that are not captured by these simple models.


What Determines the Inner Magnetospheric E?: Governing Equations

• Vasyliunas equation in MHD form (comes from neglecting inertial term in momentum equation and assuming J=0):

€

J||in − J||is

Bi− =

ˆ b ⋅r

∇V ×r

∇p

B

and the right side of (1) can be evaluated anywhere on a field line.

– The form (1) assumes isotropic pressure, but it can be generalized.

€

V = dsB∫

– where in and is mean northern and southern ionosphere, we have assumed the magnetic field strength is the same at either end of the field line,

(1)


• Ohm’s law for ionosphere:

Equations

€

Jh =t Σ ⋅ −

r ∇Φ( ) +

t Σ ⋅

r v n( ) ×

r B

Field-line-integrated current(includes both hemispheres)

Field-line-integrated Conductivity (both hemispheres)

Field-line integrals of products of Hall and Pedersen conductivities and neutral winds

• Conservation of ionospheric current:

€

r∇ ⋅

rJ h = J|| sin(I)

• Substituting (2) and (3) in (1), and neglecting winds, gives the “Fundamental equation of ionosphere-magnetosphere coupling”:

(2)

(3)

€

r∇ i ⋅

tΣ ⋅−

r∇ i( )[ ] =ˆ b ⋅

r ∇iV ×

r ∇i p sin(I )

– In the plasma sheet and storm-time ring current, the right side of this equation is typically important

• The ring current strongly affects the electric field in storm times, as Liemohn and Ridley have recently demonstrated. To calculate the electric field properly, ring-current effects must be taken into account.

• Conductance non-uniformities are also important, particularly the day-night effect.


Cartoon of I-M Coupling Equation

€

r∇ i ⋅

tΣ ⋅−

r∇ i( )[ ] =ˆ b ⋅

r ∇iV ×

r ∇i p sin(I )

Blob of plasma in a dipole magnetic field


Evolution of Magnetospheric Particle Population

where WK=kinetic energy

• For simplicity, assume that the particle distribution is isotropic– Amounts to assuming strong elastic pitch-angle scattering

• Chaotic ion motion in the plasma sheet works in this direction..

• Not really valid in inner magnetosphere.

• There is a version of the RCM (called Comprehensive Ring Current Model, CRCM), developed in collaboration with Mei-Ching Fok, that conserves the first and second adiabatic invariants, and uses different version of Vasyliunas equation.

• “Isotropic energy invariant” is conserved:

€

WK =V−2/3


Evolution of Magnetospheric Particle Population

€

∂∂t

+ rv d ⋅

r ∇

⎛ ⎝ ⎜

⎞ ⎠ ⎟η s = S − L

• Equation for evolution of the particle distribution:

where s number of particles per unit magnetic flux with certain chemical species and certain range in . S and L are sources and losses.

– s is proportional to the distribution function.

€

pV5/3 =23

η sλ s∑• Bounce-averaged drift equation:

€

rv d =

r E ×

r B

B2+

λr B ×

r ∇V −2/3

qB2

• Reference: Wolf, in Solar-Terrestrial Physics, ed. Carovillano and Forbes, 1983.


Magnetospheric Effects: Shielding• Top diagram shows equilibrium condition no

convection, with plasma-sheet edge aligned with contours of constant V.

– Particles gradient/curvature drift along contours of constant V

• Effect of applying cross-tail E (bottom) is to move edge sunward

– Causes a partial westward ring across night side

– Dusk side of edge charges +, dawn side .• Charging occurs near eq. plane and in ionosphere

– Currents flow up from dawnside ionosphere near inner edge, down to dusk side.

• Those are the region-2 currents.

• They tend to shield the near-Earth region from the dawn-dusk E. Dusk-dawn polarization E opposes convection in the inner magnetosphere.

€

J||in − J||is

Bi− =

ˆ b ⋅r

∇V ×r

∇p

B

€

rv d =

r E ×

r B

B2+

λr B ×

r ∇V −2/3

qB2


How Good is the Shielding in Steady State, for Typical Conditions?

• Answer: It’s not clear. • Shielding is often pretty good in Rice Convection Model simulations, but it is

also often marginal.• Observational uncertainty: It’s hard to distinguish steady-state magnetospheric

penetration field from neutral-wind effects. • Theoretical uncertainties:

– Because of the pressure balance inconsistency (Erickson and Wolf, 1980), it’s hard to know what value to place on pV5/3 at the tailward boundary of the RCM calculation (middle plasma sheet). I don’t think we will eliminate this uncertainty until we solve the substorm problem.

• In RCM, efficiency of shielding is sensitive to plasma-sheet temperature, with higher temperature giving weaker shielding. Reason:– The Alfvén layer lies further from Earth for more

energetic particles.– If the layer lies too far from Earth for most plasma-

sheet particles, the partial rings aren’t strong enough to shield.


Overshielding: The Idea

• If the shielding layer is configured to shield the inner magnetosphere from a strong convection field, and that convection field suddenly decreases, due to a northward turning of the IMF, the result will be a backwards E field (dusk to dawn) in the inner magnetosphere, until the shielding layer readjusts.

• Originally seen in Jicamarca data by Kelley et al. (GRL, 6, 301, 1977)– Observations were interpreted in terms

of overshielding picture


Overshielding: Pattern Detail• RCM (Spiro et al., Ann. Geophys., 6,

39, 1988) indicated that the eastward penetration field was not spread uniformly across the night side but was concentrated in the midnight-dawn sector, particularly at low L. Senior and Blanc model (JGR, 89, 261, 1984) also showed that feature.

• Agreed with earlier observations of equatorial E in response to northward turning of IMF (Fejer, in Solar-Wind Magnetosphere Coupling, 1986)

Equatorial equipotentials. Corotation not displayed.


Physical Reason for Overshielding Eastward Field Being Concentrated Midnight-Dawn

Ionospheric equipotentials for overshielding, with no distortion. Heavy arrows are E, and light arrows are Hall currents. In overshielding, low latitudes have antisunward EB drift, corresponding to sunward Hall currents. Hall currents are stronger on day side than on night side.

Hall currents remove charge from terminators, causing them to charge negative and distorting theequipotentials as shown. Dawnside equipotentialsare pushed to lower latitudes, while duskside contours are squeezed against the duskside auroral zone. Penetration to low latitudes is concentrated on dawn side.

Equatorial view of the effect. Dawnside equipotentials arePushed to the Earth, while duskside equipotentials get pushed toward the plasma sheet. Gives characteristic V-shaped equipotentials in inner magnetospehre


Recent Evidence of Overshielding From IMAGE

Data from IMAGE EUV imager. Sun is to lower right, 7:04 UT, May 24, 2000, after northward turning of IMF.

Shoulder

Shoulder

MSM simulation for same time.

From Goldstein et al., accepted for GRL, 2002


Undershielding• Undershielding is temporary penetration of dawn-dusk electric field in times

of increasing convection.• Pattern at low L is much the same as for overshielding, but the field is

reversed: westward penetration electric field post-midnight.

From Sazykin Ph.D thesis, Utah State, 2000. Equatorial ionospheric E from Fejer and Scherliess (JGR, 102, 24047, 1997).

• The main physical reason is the same as the one given for the concentration of eastward E field in that sector in overshielding


Ring Current Injection

• The injection of the ring current in the main phase of a major magnetic storm involves massive undershielding.

• A westward electric field on the night side injects the ring current deep into the magnetosphere.

• The new ring current forms a partial ring early in the main phase, but forms a complete ring eventually.

• According to the conventional wisdom, the partial ring is centered near local dusk, because that is where the ion Alfvén layer comes closest to Earth if the convection field is dawn-dusk.


Ring Current Injection: Conflicting Cartoons• The magnetic field decrease observed in the early

main phase at low latitudes on the Earth’s surface has a clear dawn-dusk asymmetry: the depression is much greater on the dusk side. – Conventional wisdom associates this with an

asymmetric ring current, centered near dusk.

• Picture of region-2 currents and shielding, which had the plasma sheet coming closest to Earth near local midnight.– Suggests partial ring centered near midnight.

– The dawn-dusk asymmetry in the ground magnetic signature was interpreted in terms of the sum of region-1 and region-2 currents being down on the day side, up on the night side, with connection through antisunward Hall currents in the auroral zone (Wolf et al., JGR, 86, 2242, 1981; Crooker and Siscoe, JGR, 86, 11201, 1981; Chen et al., JGR, 87, 6137,1982)


IMAGE Observations of Ring Current Injection

• Observations and CRCM model fluxes for 12 August 2000, at the peak of the main phase of a storm. Ring current peaks between midnight and dawn in both observation and model. From Fok et al. (submitted to Space Sci. Rev., 2002)

IMAGE ENA, 27-39 keV CRCM Model, 8 UT, 32 keV


CRCM Equipotentials

• In the spirit of full disclosure:– Most RCM runs show pressure peaks near local midnight in ring-current injection.

The degree of potential twisting and the location of the pressure peak vary among different RCM storm runs and we haven’t figured out what controls them. Some possibilities:

• strength of convection• plasma sheet ion temperature• background conductance (including sunspot number and dipole tilt)• auroral conductance enhancement

Note the extreme twisting of the CRCM equipotentials, with the westward electric field centered near dawn. From Fok et al. (submitted to Space Sci. Rev., 2002)


Effect of Strong Penetration on Equatorial Ionosphere

• Basu et al. (GRL, 28, 3577, 2001) showed dropout of the equatorial ionosphere at 840 km altitude in main phase of the Bastille Day storm. They interpreted that as the result of upward drift of the F-layer above that altitude.

• This uplift was accompanied by strong scintillations.

– Eastward electric field causes a downward drag that acts like increased gravity and encourages the Rayleigh-Taylor instability that causes spread F.

• Note: Massive rearrangements of the low-latitude ionosphere imply massive changes in conductance. Proper modeling of something like this requires active coupling of ionosphere and magnetosphere models–haven’t done that yet.


RCM Simulation of Bastille Day Storm

• Source: Stanislav Sazykin• Movie shows potential in

equatorial plane– Top panel shows polar-cap

potential (blue) and Dst (white)– This simulation was done with

IRI model of sunlight-driven ionosphere, with auroral enhancement

• Notice:– Skewing of equipotentials– Strong westward flow at low L

from dusk to past midnight (Subauroral Polarization Stream, discussed shortly).

– Outward flow on dusk side near Earth

Movie is file 0700_Veq_iri_short.avi


RCM for Bastille Day Storm– Plots of Ionospheric E

• At low latitude, strong westward E in post-midnight sector, eastward E across night side.

• Snapshot shows an example of local time dependence of ionospheric electric fields, for latitudes of 15°, 40°, and 55° at 22:20 UT on 15 July, late in the main phase.


Bastille Day with SUPIM Conductance

Model• Plot shows E fields for 1240 UT on July

15.

• This conductance model has sharper conductance jump at terminator. Note eastward E concentrated post-dusk.

• Overall conclusion:– A huge storm like this can result in

strong eastward electric fields on the dusk side.

• The local-time distribution of this eastward field depends strongly on the details of the ionospheric model.

• To treat this properly, really need a model with an active ionosphere, because conductance is clearly affected by the dramatic magnetosphere-caused layer motions. The conductance acts back on the electric field.


SubAuroral Polarization Streams (SAPS)• RCM simulations of active conditions

frequently show strong flows in the subauroral region

• Example–Trevor Garner’s RCM simulation of June, 1991 storm

– Rapid flow is just earthward and equatorward of the electron plasma sheet

– That flow occurs in the dusk-midnight sector, sometimes extending past midnight.

– Note that there are stronger electric fields in the inner magnetosphere than in the tail – the opposite of shielding.

– This feature has been observed in strong storms by CRRES (Rowland&Wygant, JGR, 103, 14959, 1998), Burke et al. (JGR, 103, 29399, 1998) (see bottom panel).

– Also observed very convincingly from Millstone Hill (Foster and Burke, subm. EOS). They coined the name.

(Garner, Ph.D. Thesis, Rice Univ., 2000)


Physical Interpretation of SAPS

• In the pre-midnight sector, plasma-sheet ions penetrate closer to Earth than electrons.

• Electrons mostly control ionospheric conductance.

• Therefore, in the premidnight sector, the inner edge of the plasma-sheet ions lies at lower L than the auroral conductance enhancement.

• Most of the shielding (region-2) current is driven by ions, because they carry most of the pressure.

• Therefore, some region-2 current flows into low-conductance, subauroral ionospheric region in the pre-midnight sector. That is what causes SAPS in the RCM.

Electrons

Ions



• The obvious interpretation of SAPS is that they are the strong E field region between inner edge of region 2 and the equatorward edge of the auroral electrons.

• This gap between these two regions becomes larger in great storms, because precipitation erodes the electron inner edge.

• A peak in sunward flow velocity just equatorward of the duskside auroral is a usual feature of RCM simulations for active conditions.



• This is essentially the same interpretation as Southwood and Wolf (JGR, 83, 5227, 1978) for SubAuroral Ion Drift events (SAIDs).

• SAID events (also called polarization jets) were discovered by Galperin (Kosm.Issled.,11,273,1973) and by Spiro et al. (JGR, 83, 4255, 1978).– SAID events are narrow (~1º wide) and occur frequently.

• Banks and Yasuhara (GRL, 5, 1047, 1978) suggested a mechanism involving conductivity reduction due to the very fast drift. This was essentially an ionospheric instability.

• Both types of mechanisms probably operate. Foster has suggested that the broad fast-flow region (SAPS) and the superposed narrow features (SAID) are different phenomea. That is an appealing interpretation: SAPS are essentially a magnetospheric phenomenon and SAIDS are due to ionospheric instability.


Summary• Rice Convection Model and similar models that self-consistently

calculate inner-magnetospheric electric fields and display features that agree with observed features, some of which only recently have been clearly identified. – Overshielding: equatorial radar, new IMAGE plasmapause measurements– Undershielding: equatorial radar, IMAGE ring-current injections– SAPS and SAIDs: magnetospheric and ionospheric E fields.

• Models are less good at getting quantitatively accurate electric fields at any given time.

• Many details are not worked out: what controls degree of skewing of undershielded penetration field in large storms, relationship between SAPS and SAIDs. We can’t reliably calculate how good shielding should be in steady state.

• We really need a coupled global-magnetosphere/ring-current/thermosphere/ionosphere model. Such big coupled models should result from the big Michigan and Boston University modeling efforts.

Date post:	27-Dec-2015
Category:	Documents
Upload:	leon-cox
View:	219 times
Download:	1 times

Tutorial on Inner Magnetospheric Electric Fields, GEM Workshop, 6/021 Inner Magnetospheric Electric...

Documents