IKI, Feb. 20071

Новые сведения о структуре и динамике магнитосферных

токовых слоев

В. А. Сергеев (СПбГУ),

Thanks to : S. Dubyagin, D. Sormakov, S.Apatenkov (SPbU), Y.Asano, A. Runov, W. Baumjohann, R. Nakamura, T. Zhang (IWF), A.Petrukovich (IKI), P. Louarn , J.-A. Sauvaud (CESR), A.Fazakerley, C.Owen (MSSL), T. Nagai (TITECH), T. Mukai (ISAS)

INTAS grant 03-51-3738

IKI, Feb.2007 2

Предмет исследований

Токовый слой хвоста и МП – наиболее досягаемы для прямых измерений, Cluster, Geotail

Задачи гранта INTAS (2003) : • Свойства и происхождение динамичных токовых слоев (flapping motions)• Структура слоев• Особенности ТС вблизи области пересоединения

До этих работ – ТС представлялся преимущ. плоским образованием (~XY GSM), Основная физ.модель – слой Харриса (Bn=0), обобщенная на случай Bn>0 при сохранении 1d аппроксимации

Подавляющая часть приведенных здесь результатов получена при быстрых пересечениях (flapping) токового слоя Cluster, Geotail

Geotail

Cluster

Bx

IKI, Feb.2007 3

I. Характеристики ‘flapping’ токовых слоев

IKI, Feb.2007 4

Data base of rapid CS crossings / Methods

CLUSTER r ~1800,1000, 300kmFGM + CIS data 2001, 2004, 2003

Data set #1 (186+189 events): Jump Bx : |Bx | > 15 nT Duration: 30s< τ < 300 s No special selection of plasma properties.

Data set #2 (57 events, reliable normals) Baricenter cross the neutral sheet; Smooth crossings, similar traces at all SC, stable timing; Good curlometer results (DivB/Curl B < 0.3 in > 60% of points);

Tilt determination 1/ MinVarAnalysis (1SC) n1 , n2 (~j), n3 (~n) in 1d CS2/ Current-based (4SC): nj = [n1 x jNS] / | n1 x jNS |3/ Timing (4SC)

[Runov et al, AnnGeo, p.1391, 2005]:

B1 n

j

Bx

IKI, Feb.2007 5

Methods : How accurate are tilt determinations ??

No a priori reason why MVA should work in 1D current sheets..

Comparison of CS normal directions determined with timing (TIM), current (CUR) and MVA are in good agreement : (n1*n2)>0.9 < 25o

for >80% cases TIM/CUR, MVA is less accurate

> Confirm local planar approximation as reasonable starting choice.

[ Sergeev et al., Ann.Geophys.,p. , 2006]

100%

Practical criteria for MVA in tail CS :

2 / 3 > 4 : ~25o, (incl ~75% events)

IKI, Feb.2007 6

Example : October 20, 2001 CL at [-13, 12,-1] Re

Multiple NS crossings during substorm (onset ~0920UT), slow flow Vx,y (passive CS), all time inside PS (Sergeev et al.,2004 GRL)

Consistence of n determinations from T and MVA – very large tilt toward Y !!

Consistent alternating nz (see Table) Alternating signs jz (+B,-b, etc) and nz variations under positive jy; |jz| |jy| – large tilts

Propagate duskward

Large tilts, YZ Kink mode with strong folds

Z

Y j , n

NS

IKI, Feb.2007 7

20040805 Cluster & TC-1, Kink mode confirmed

8 NS crossings, timescale ~150 – 500 s , weak substorms Similar ~phased variations of Bx (1320-15 UT) Cluster & TC-1 kinks with length >5 Re

along the tail

[Zhang et al., Ann Geo, 2005]

TC-1: [-10.7 -6.8 2.6]Cluster: [-15.4 -8.9 3.4] Cluster Timing & MVA: Normals swinging in YZ plane Kinks propagating dawnward at V~30 km/s

CL

TC-1

YXZ

Tail

Dusk

Dawn

IKI, Feb.2007 8

Appearance of Flapping Structures

SHAPE (reconstructed from probing small parts !):

appear as parts of oscillation pattern or solitary structures ; parts of plane not uncommon

Y AZ h CS ?1-3Re 0.5-2Re 0.2-1Re Large tilts of CS parts (up to 90o to nominal Z)

Large-amplitude Nonlinear Structures (Waves)

LIFETIME : Only lower estimates are possible in examples with slow propagation which give

TL can be 5 min

5 min

IKI, Feb.2007 9

Guide- versus normal B-field components

Tilted sheets possess large guide B-field as distinct from sheets with standard tilt.

B vector is averaged over |BX |<+3nT region.

α - angle between n and Z GSM

0 10 20 30 40 50 60 70 80 90

, deg.

0.01

0.1

1

10

100

| Bno

rm /

Bgu

ide |

0.01

0.1

1

10

100

H orizonta l sheets (standard)

Vertica l sheets

< |B guide|>= 3.5 nT< |B norm |>= 1 .4 nT

<|B guide|>= 10.4 nT<|B norm |>= 1.7 nT

B1 n

j

Standard Tilted CS

BJ/Bn

[ Dubyagin et al., AnnGeo, 2007 subm ]

IKI, Feb.2007 10

Kink Propagation and Ion flows along normal direction

ION FLOWS (from CIS) : Problems: effect of same order of magnitude as possible instrumental Z-offsets and errors; transient effects and changing tilts easily ruin any correlation; only Vn is available from timing

Special subset : 42 events with nTnJnMVA within 20o

RESULTS :

Plasma flows in the same direction with ~same velocity along the normal;

Most of flapping Kink structures – can be the structures nearly standing in the flow

[ Dubyagin et al., AnnGeo, 2007 subm ]

0 20 40 60 80 100 120V n p lasm a , km /s

0

20

40

60

80

100

120

Vn f

ront

, km

/s

r = 0.74N = 42

2001 - C IS -C O D IF - C L42004 - H IA - C L1 + C L3

H IA m om ents during S torm tim eare excluded

IKI, Feb.2007 11

Timing : propagation direction & velocities

Propagation from center to flanks, perpendicular to ‘magnetic field plane’;

Zone of ‘mixed direction‘ near midnight (esp. premidnight) where X-line observations are most frequent at substorm onset (Nagai et 1998);

Propagation speeds 20-300 km/s, median ~ 50 km/s;

Larger speeds near MReconnection site and in BBFs;

Propagation pattern reminds the pattern of cross-tail average flow, could be structures standing in the flow???

VT

B1

Runov et al., AG 2005

Sergeev et al., GRL 2004

IKI, Feb.2007 12

Occurrence of Fast CS crossings

Geotail : 3 years data set, 3s B-field from DARTS; 1100 fast crossings (occurrence); MVA with 2/3>4 and N>15 to get normals (480 events, for tilts study);

Results n3 [~0, nY, nZ] like at Cluster confirm YZ-kink mode

duration depends on local flows (short during fast flows and in reconnection regions)

[Sergeev et al. AnnGeophys, 2006]

GT CL

MVA normals

Dawn Dusk

IKI, Feb.2007 13

Occurrence of Fast CS crossings, 2

Results n3 [~0, nY, nZ] confirm kink mode

duration - short during local fast flows)

occurrence frequency increases downtail and in Y ~ [-5..+10] Re

consistent with flankward propagation, confirm magnetospheric origin

occurrence resemble occurrence of BBFs and MR at substorm onsets (Nagai et al.,1998)

[Sergeev et al. AnnGeophys, 2006]

IKI, Feb.2007 14

Flapping : Activity Dependence

Dependence on AE and auroral activity Common during substorms, but sometimes met during quiet AE Usually associated with auroral brightenings or poleward arcs (PC or streamer?), but clear one-to-one relationship occur rarely few events with NO auroral brightenings (all – CDPS events).

Dependence on Local activity flapping appear within the whole range of PS parameters sharp crossings with large tilts and large Vz and VT are frequently noticed near MR region, and in BBFs (?);

Needs more (quantitative) study[ Dubyagin et al., paper IAGA2005-A-00302 ]

IKI, Feb.2007 15

duration <30 - >300s, depends on local flows (short in BBFs and near MR) in the neutral sheet Bns ~4 nT (1..18nT), usually |Bj*|>Bn* curvature –Earthward (closed FL-~90%) in CL data set; B-curvature radius Rc ~ 5 Lcp (Lcp -gyroradius in BL field) halfthickness h~10 Lcp (1..20 Lcp), weak correlation with Rc adiabaticity parameter = {Rc BL/ (Lcp Bns)}1/2 ~0.6 (0.2..2) - non-adiabatic ions any local (density/temperature/velocity) conditions plasma velocity: BBFs only in ~25%, >60% Vp<100km/s

Parameters of flapping current sheets

Runov et al. AnnGeo, 2005, p.1391, also

Shen et al JGR, 2003, p.1168

IKI, Feb.2007 16

Large-amplitude Nonlinear Structures (Waves or Solitary structures); Very large tilts (>45o from Z in ~50% of events) Fast CS crossings due to kink ‘waves’ with normals rotated B (in ~YZ); Propagated ‘centerflanks’ , ion speeds 20-300 km/s, median ~50km/s; Occur 10->30 Re, frequency increases with distance, peaks premidnight Associated with auroral brightenings and substorms, exceptions exist Ion flow includes :

Vz corresponding to up/down CS motions; Vpn flow have same sign/magnitude as the propagation velocity for low speed events (?);

Summary of Flapping-related structures

IKI, Feb.2007 17

Mechanisms of Flapping ?

MHD effect of localized reconnection with magnetic shear (Semenov et al.,1994, Shirataki, Fujimoto et al 2006) Kinetic (KH or drift) instability (Nakagawa-Nishida, 1989, Karimabadi et al., 2003,…. ) Standing modes in (neutral Harris) sheet (Fruit et 2002, Volverk et al., 2003) Eigenmodes in finite Bz current sheet (Golovchanskaya, Maltsev, 2004) PIC simulations of CS instability (e.g. with real hcs, Sitnov et al 2004), association with other instabilities (e.g. LHDI, etc)

Different initial models and parameters, different detailness of predictions - difficult to compare ;

critical issues to answer :

MHD or kinetic effects? Instability or generation by reconnection? - ???

Propagation velocity (in dHT frame) ? -typically small CS parameters/structures favourable for generation? - no obvious preference..

Different mechanisms could be involved???

IKI, Feb.2007 18

II. Структура ‘flapping’ токовых слоев

IKI, Feb.2007 19

October 20, 2001: Oscillation amplitude and CS structure

Vz integration – vertical PS/plasma motions: consistent variations Vc & Vp ; z~1-2 Re

CS structure of individual crossings, from B (z*) (Vc and average tilts used) :

different shapes (single peaks, flat, bifurcated, symm/asym, time-dependent?,…) ;

CS scale-sizes vary between ~1000 km and ~4000 km;

All crossings are different !

A B b C c D d E e

Thin embedded,

Bifurcated,

1 asymmetric, etc

IKI, Feb.2007 20

Survey of structures , comparison to Harris distributions

Reconstruction of 30 suitable crossings (tilts, coord.system, Z*scales); Identify :

central embedded sheets ; bifurcated sheets; asymmetric sheets

Non-Harris distributions !! j shape deviate from Harris; Density and j- profiles are different, flat density distributions

(Runov et al., AnnGeo 2006)

IKI, Feb.2007 21

Parameters of Current sheet structures

~55%: Central peak, main part of the current is concentrated in |Bx|<0.5BL, BL=(B2+2μ0Pi)½

Typical half-thickness of the central peak ~ 1000 - 2000 km ~ 3 – 7 Lcp, “Shoulders”

~10%: Bifurcated: Local minimum of the current density at Z* = 0 (Bx~0) Typical half-thickness ~ 3000 km ~ 10 LcpThicknesses of peaks of 1000 - 2000 km

~35%: Asymmetric: Off-equatorial maximum of the current densityTypical half-thickness ~ 1500 - 3000 km No big difference in parameters !

Also - No significant differences in ion anisotropy/ nongyrotropy

IKI, Feb.2007 22

Current sheet structures

Are non-Harris features special for flapping sheets only (e.g. transient effect??)

Asano et al 2005 (GRL, L03108):- Other approach and method (no preference to flapping sheets);- Large tilts excluded;

Results: - Harris distributions – rare !- Central embedded sheets - ~40%- Bifurcated/off-center peaks - ~17%comparable to % in flapping sheet survey!

-% of bifurcated sheets slightly increases during BBFs

- similar conclusion by Thompson, Kivelson et JGR2006

B

B

Harris

IKI, Feb.2007 23

Inconsistence of rotB and ion currents

Expected: Ti/Te>>1 jp>>je and main contribution from jp to jTOT

Observed: large discrepancies : different signs jC and jp<0, |jp| > jC Implies electrons as main current carriers ?

Explanation – dawnward convection, implies converging En and negative e-charge in the sheet :

(Runov et al., AnnGeo 2006)

jC = n e (<Vp> - <Ve>)

IKI, Feb.2007 24

Comparisons including electron contribution

Calibration of electron spectrometer at required accuracy appeared to be very difficult instrumental problem

Results are yet available only for 1 long event in 2001,

Seem to be consistent with

electrons to be dominant current carrier..

B1, nT

IKI, Feb.2007 25

Recent findings (CLUSTER):

Hall B-fields and current/carriers; mostly following predictions;

Thin sheets ~1 c/pi (400-800km) directly confirmed, j ~ 30-50 (up to 120) nA/m2 ;

may be very short (seconds) large tilts are common (!?) may include small FRope-like structures etc,

entire picture sometimes looks like strong turbulence…

Nakamura et al., JGR , A11206, 2006, Wygant et al., JGR, 2004

CS in/near Reconnection Region

10secCSheet in ~XZ gsm plane

~jzgsm

flapp

ing-

like?

X~-15Re

IKI, Feb.2007 26

IV. Что мы узнали?

IKI, Feb.2007 27

Discussion: Z-структура ТС – 1d or 2d effect?

До этих работ – ТС представлялся плоским образованием, 1d слой Харриса (Bn=0), обобщенный на случай Bn>0

Основой 1d гипотезы обычно считается /z >>/x ; однако при Bx>>Bn 0 X>>Z (‘Tail approximation’) распределения Р(z), j(z) определяются их Х-распределениями

XВ эксперименте (преимущ. flapping ТС) : -структура\масштаб последовательных слоев изменчивы (20\10\01, Nakamura et 2006), -возможна NS-асимметрия (t -? или структура-?), -вероятны уединенные структуры

>Каков возможный критерий идентификации 1-2 мерности наблюдаемых слоев ?

IKI, Feb.2007 28

Discussion: Z-структура ТС – 1d or 2d effect? (2)

В эксперименте (преимущ. flapping ТС) : -структура\масштаб последовательных слоев изменчивы (20\10\01, Nakamura et 2006), слабо связаны с Z- масштабом структур-возможна NS-асимметрия (t -? или структура-?), -вероятны уединенные структуры

>Каков возможный критерий идентификации 1-2 мерности наблюдаемых слоев ?

3D MГД моделирование: - YZ-kink – встречаются часто;- частое сосуществование структур разного типа (по X и Y!!); нередко Y ~ Z, толстые слои - быстрые изменения во времени;

МГД-система производит многообразие структурпреимущ. за счет 2d (+нестац.) эффектов магнитного пересоединения

См.также [Thompson, Kivelson et al., JGR 2006]

IKI, Feb.2007 29

Future work

IKI, Feb.2007 30

Curlometer and proton current comparisons

Explanation – requires dawnward convection (~50 km/s), implies converging En (+ earthward E in NS) and negative e-charge in the sheet :

origin of dawnward convection on duskside??? – inconsistent with average convection

consistent with Wygant et al JGR 2005 Cluster event analyses but obtained in thin electron-scale CS, charge separation near MR as compared to relatively thick (h~few Lcp) events in our survey (?) ;

A challenge:- to confirm this explanation by studying electron contributions and - to understand origin/mechanisms of dawnward convection in the sheet;

jC = n e (Vp-Ve) YZ plane

Ez* Vy*

Hori et al.,2000, also

Kauffmann et 2001, Asano et 2003 etc