Features of Pc5 Pulsations in the Geomagnetic Field ... · GEOMAGNETISM AND AERONOMY Vol. 56 No. 1...

ISSN 0016�7932, Geomagnetism and Aeronomy, 2016, Vol. 56, No. 1, pp. 42–58. © Pleiades Publishing, Ltd., 2016.Original Russian Text © V.B. Belakhovsky, V.A. Pilipenko, S.N. Samsonov, D. Lorentsen, 2016, published in Geomagnetizm i Aeronomiya, 2016, Vol. 56, No. 1, pp. 46–63.

42

1. INTRODUCTION

Geomagnetic Pc5 pulsations (periods of approxi�mately several minutes) are among the most powerfulmanifestations of MHD activity in near�Earth space.MHD waves are generated by the energy that is sup�plied from the interaction between the solar wind andthe magnetospheric boundary or inequilibrium fluxesof energetic particles in the magnetosphere. ExcitedPc5 pulsations can reach considerable amplitudes (upto several dozen nT within the magnetosphere) owingto a resonance enhancement in the Alfvén resonanceregion.

Pc5 pulsations can pronouncedly modulate fluxesof trapped energetic electrons and different�energyprotons in the magnetosphere (Sarris et al., 2007), aswell as the ionospheric plasma (Pilipenko et al., 2014).In this case the modulation depth of the magneto�spheric particle and ionospheric plasma densities canpronouncedly exceed the geomagnetic field modula�tion depth. The modulation of precipitating energetic

electron fluxes (~30–100 keV) by Pc5 pulsations man�ifests itself during ground�based observations in theappearance of periodic riometer absorption pulsations(Olson et al., 1980; Belakhovsky and Pilipenko, 2010).

The interaction between MHD waves and mag�netospheric particles can result in pulsations of theforms and intensities of different�type auroras. Modu�lation of the auroral luminosity by different ULF geo�magnetic pulsations was observed in different localtime sectors, which can indicate that modulation pro�cesses of auroral electron fluxes (~1 keV) are different.Wave�like disturbances of diffuse auroral luminosity(undulation), which were caused by direct plasmasheet electron precipitation, were observed (Baishevet al., 2012). Discrete auroral arcs can also pulsatewith the frequency of Pc5 geomagnetic oscillations(Vorobjev et al., 2008; Roldugin and Roldugin, 2008).In this case an arc initially appears at the auroral ovalequatorward boundary, from which it moves polewardat a velocity of ~1 km/s and subsequently fades.

Features of Pc5 Pulsations in the Geomagnetic Field,Auroral Luminosity, and Riometer Absorption

V. B. Belakhovskya, V. A. Pilipenkob, S. N. Samsonovc, and D. Lorentsend

a Polar Geophysical Institute, Apatity Division, Kola Scientific Center, Russian Academy of Sciences,ul. Fersmana 14, Apatity, Murmansk oblast, 184209 Russia

b Institute of Physics of the Earth, Russian Academy of Sciences,Bol’shaya Gruzinskaya ul. 10, Moscow, 123995 Russia

c Institute of Cosmophysical Research and Aeronomy, Siberian Branch, Russian Academy of Sciences,pr. Lenina 31, Yakutsk, 677007 Russia

d University Center on Spitsbergen, Norwaye�mail: [email protected]

Received April 21, 2015

Abstract—Simultaneous morning Pc5 pulsations (f ~ 3–5 mHz) in the geomagnetic field, aurora intensities(in the 557.7 and 630.0 nm oxygen emissions and the 471.0 nm nitrogen emission), and riometer absorption,were studied based on the CARISMA, CANMOS, and NORSTAR network data for the event of January 1,2000. According to the GOES�8 satellite observations, these Pc5 geomagnetic pulsations are observed asincompressible Alfvén waves with toroidal polarization in the magnetosphere. Although the Pc5 pulsation fre�quencies in auroras, the geomagnetic field, and riometer absorption are close to one another, stable phaserelationships are not observed between them. Far from all trains of geomagnetic Pc5 pulsations are accompa�nied by corresponding auroral pulsations; consequently, geomagnetic pulsations are primary with respect toauroral pulsations. Both geomagnetic and auroral pulsations propagate poleward, and the frequencydecreases with increasing geomagnetic latitude. When auroral Pc5 pulsations appear, the ratio of the557.7/630.0 nm emission intensity sharply increases, which indicates that auroral pulsations result from notsimply modulated particle precipitation but also an additional periodic acceleration of auroral electrons bythe wave field. A high correlation is not observed between Pc5 pulsations in auroras and the riometer absorp�tion, which indicates that these pulsations have a common source but different generation mechanisms.Auroral luminosity modulation is supposedly related to the interaction between Alfvén waves and the regionwith the field�aligned potential drop above the auroral ionosphere, and riometer absorption modulation iscaused by the scattering of energetic electrons by VLF noise pulsations.

DOI: 10.1134/S001679321506002X

GEOMAGNETISM AND AERONOMY Vol. 56 No. 1 2016

FEATURES OF Pc5 PULSATIONS IN THE GEOMAGNETIC FIELD 43

According to the Polar satellite observations of theauroral oval UV luminosity, the effect of modulationby Pi3 pulsations, which are generated when substormdevelops, were detected in a wide local time intervalfrom ~1700 to ~0300 LT (Liou and Takahashi, 2013).Modulation of diffuse auroras by Pc5 pulsation in thedawn sector was observed as local luminosity bright�ening (auroral patch), which was synchronized withgeomagnetic pulsations and moved poleward andwestward at velocities of ~1 and ~15 km/s, respec�tively. Luminosity variations looked like successive“on”–“off” switchings (Yamamoto et al., 1988).

More and more experimental evidence for such asubstorm development scenario has recently appeared.According to such a scenario, the initial instabilitydevelops in the near�Earth region of closed field lines.The unbalanced magnetospheric system subsequentlyreleases the main energy as a result of reconnection inthe tail. None of the numerous, theoretically possiblemechanisms can be considered closely reasoned orrejected as a priori unreal. In some theoretical models,Alfvén oscillations of field lines (with toroidal orpoloidal transverse polarization) and their activeinteraction with the ionosphere are of fundamentalimportance (Samson et al., 1992; Rae et al., 2014).

In addition, Pc5 waves can not only passively mod�ulate magnetospheric particles but also result in theiracceleration. However, a specific acceleration mecha�nism has not yet been determined, even though manytheoretical models were proposed. The periodic accel�eration and precipitation of auroral electrons are sup�posedly caused by a field�aligned electric wave field(E||). Electrons accelerated by a wave field can receiveenergy sufficient for aurora brightening on the whole;i.e., they can result in the formation of the so�called“Alfvén aurora” (Keiling et al., 2002). It is possible todifferentiate between Alfvén aurora and aurorascaused by a stationary electron acceleration in a regionwith a longitudinal potential drop by only the differ�ences in the spectrum of auroral electrons registeredon low�orbiting satellites.

Detailed studies of simultaneous periodic pulsa�tions of auroras and the geomagnetic field were per�formed rather rarely, since such effects can beobserved only when several favorable factors coincide.Therefore, each registered event is of interest. Simul�taneous high�latitude Pc5 pulsations in the geomag�netic field, auroras, and riometer absorption werestudied in the event analyzed by Rae et al. (2007).These authors detected the effect of auroral pulsationpoleward propagation. However, the observed patternwas complicated by the superposition of magneto�spheric wave processes with different frequencies(~1.8 and ~3.0 mHz). Auroral Pc5 pulsations wereinsufficiently distinct, and the relations between thesepulsations were not analyzed in detail. A shift by ~π/2between magnetic and riometer variations remainedunclear.

Daytime pulsating (with a periodicity of 2–15 min)auroral arcs propagating poleward (poleward movingauroral forms, PMAFs) are among the auroral phe�nomena related to ULF geomagnetic oscillations.These narrow arcs with a width of ~50–100 km extendazimuthally over ~500–1000 km and move polewardat a velocity of ~1.5 km/s. PMAFs are observed in thehigh�latitude ionosphere in the dayside sector(~0900–1500 MLT). Reconnection at the daysidemagnetopause was considered among the PMAFsources (Oieroset et al., 1997). However, a relationwas detected between the PMAF appearance in thevicinity of the cusp near the convection reversalboundary and Pc4 oscillations (~2 min) of the geo�magnetic field (Milan et al., 1999). The frequency ofoptical pulsations was approximately twice as high asthat of geomagnetic pulsations. The azimuthal propa�gation characteristics differed according to the optical(the wave number is m ~ 20–30) and magnetic (m ~ 9)data. On the whole, most observers assumed that thepoleward motion of auroral arcs is related to the phasestructure of the geomagnetic pulsation field.

Simultaneous observations of Pc5 pulsations inauroras and riometer absorption could provide impor�tant information about the character of the interactionbetween waves and particles in the magnetosphere,since these pulsations are caused by electrons with differ�ent energies. Data on the intensity of the red (630.0 nm)

and green (557.7 nm) oxygen and nitrogen ( ) line(471.0 nm) emissions are most widely used. The greenoxygen emission is caused by suprathermal electronswith energies about several keV; red line emission iscaused by thermal electrons with energies of severalhundred eV; nitrogen emission is caused by electronswith energies of approximately several keV. The char�acteristic lifetime of the excited state of oxygen is ~1 sand ~2.5 min for the green and red emissions, respec�tively. Ionization of the lower ionospheric layers,which results in an additional absorption of cosmicradio noise registered with riometers, is caused byelectrons with Е ~ 30–100 keV.

The present work analyzed an event that makes itpossible to obtain important information about themechanism by which auroral arcs are modulated in thedawn sector and about their relation to correspondingPc5 pulsations in the geomagnetic field and riometerabsorption of space radio noise.

2. GROUND�BASED AND SATELLITE OBSERVATIONS

The 1�min data of GILL, PINA, and FSMI merid�ional scanning photometers from the NORSTAR net�work (Canada) were used in auroral observations. Thecoordinates of these stations are given in the table. TheNORSTAR photometers register auroras in the 557.7,630.0, 471.0, and 486.0 nm emissions and give data for17 angles. It is assumed that the red (630.0 nm) emis�

N2+

44


BELAKHOVSKY et al.

sion maximum is located at an altitude of 230 km; thegreen (557.7 nm) emission maximum is at 110 km alti�tude. Data from the all�sky camera and scanning pho�tometer from the RANK high�latitude station wereabsent for the considered event.

Data from the NORSTAR vertical riometers (seetable) along the ~330° geomagnetic meridian wereused to control the precipitation of energetic (~30–100 keV) electrons. Data from the ground stationsfrom the CARISMA and CANMOS networks (the

schematic location of these stations is presented inFig. 1) were used to observe geomagnetic field varia�tions. These stations form the latitudinal profile alongΛ ~ 330° geomagnetic longitude and the longitudinalprofile at a geomagnetic latitude of Φ ~ 65°–66°.

The ground�based observations were completedwith the data of three�component magnetometers onthe GOES�8 (MLT = UT – 5) and GOES�10 (MLT =UT – 9) geostationary satellites. The orientation ofthe magnetic components is as follows: the radial

Ground stations

Station IAGA code Geomagneticlatitude, °

Geomagnetic longitude, ° MLT Observation type

Rankin Inlet RANK 72.22 335.9 UT – 6.2 М, R

Eskimo Point ESKI 70.52 333.2 UT – 6.4 М, R

Fort Churchill FCHU 68.32 333.5 UT – 6.4 M

Fort Smith FSMI 67.28 306.9 UT – 8.1 М, R, P

Rabbit Lake RABB 66.85 319.1 UT – 7.3 М, R

Gilliam GILL 66.03 333.1 UT – 6.4 М, R, P

Poste de�la�Baleine PBQ 65.02 352.5 UT – 5.1 М

Fort McMurray MCMU 64.17 309.2 UT – 8.0 М, R

Island Lake ISLL 63.62 333.4 UT – 6.4 М, R

Pinawa PINA 59.98 331.8 UT – 6.5 М, R, P

(M) Magnetometer, (R) riometer, and (P) photometer

70

65

–12060

55

50

–100

–80

–140

LANL–1989

DAWS

GOES–10

FSMI

MCMU

RABB

RANK

ESKI

FCHU

GILL

ISLL

PINA

PBO

GOES–8

TALO

Fig. 1. Location of ground magnetic stations (triangles) at the CARISMA and CANMOS networks, photometers (squares), andriometers (asterisks). The projections of field lines passing through the geostationary satellites are marked by filled circles.



component (Be, toward the Earth); perpendicular tothe ecliptic plane (Bp); azimuthal (Bn, normal to thefirst two components). Figure 1 shows the projectionof the field lines, passing through these geostationarysatellites, onto the ionosphere (GOES�10 (64°, 298°),GOES�8 (64°, 357°), and LANL�1989 (64°, 268°)).

We also considered the data of the energetic parti�cle detectors on the LANL geostationary satellites.LANL�89 (MLT = UT – 11), the closest satellite tothe 330° profile, which was shifted by approximately4.5 h (~65°) westward according to MLT. The solarwind and IMF parameters were taken from the OMNIdatabase and the ACE satellite. All data used in thework were recalculated for the 1�min time interval.

Sixteen distinct events with simultaneous pulsa�tions of the auroral luminosity, geomagnetic field, andriometer absorption according to the GILL stationdata were detected in 2000–2009. Only one of charac�teristic events (on January 1, 2000) is presented in thework.

3. EVENT OF JANUARY 1, 2000

3.1. Geophysical conditions and interplanetarymedium parameters. The analyzed Pc5 geomagneticpulsations were observed at 1100–1600 UT during thesubstorm (AE up to ~1100 nT) (Fig. 2) in the recov�ery phase of a weak magnetic storm (with the inten�sity SYM�H ~ –60 nT) that started on December 30,1999. The SYM�H index varied from approximately–20 to –40 nT when Pc5 geomagnetic pulsationsappeared. The solar wind velocity was rather high(~700–750 km/s) at a low density (~2–4 cm–3); pro�nounced jumps in the solar wind velocity were notobserved in this case. A geomagnetic field bay causedby substorm was especially pronounced on GILL(~66°): the field value started decreasing at ~1200 UTand reached its maximum at ~1300 UT (Fig. 2).

Particle detectors on LANL�89 registered theinjection of electrons with energies varying from 50 to500 keV, coincided with a magnetic bay at ground sta�tions. The electron fluxes started increasing at~1200 UT and reached their maximal values at 1230–1300 UT for E = 50–150 keV. The increase in the fluxintensity had a dispersion related to energies: lowenergies started increasing earlier than high energies.The injection of energetic electrons is accompanied byenhancement of the riometer absorption and auroralluminosity (Fig. 2). The observations of auroral emis�sions continued to ~1340 UT and subsequently endeddue to the daytime illumination.

3.2. Pc5 geomagnetic pulsations on the ground.Intense Pc5 pulsations with an amplitude reaching~300 nT for the Х (Fig. 3) and Y (not shown) componentswere observed in the dawn sector (~0500–0800 MLT) at1100–1600 UT on the meridional profile of magneticstations along ~330° longitude at latitudes of ~70°(ESKI) to 60° (PINA) (Fig. 3).

The pulsation spectral composition (Fig. 4)includes several characteristic frequencies and varies dur�ing the substorm. The spectral maximum at ~5.0 mHz(Fig. 4a) predominated at all latitudes in both compo�nents from 1120 to 1240 UT, up to the instant whenthe substorm intensity became maximal. The spectralpower along latitudinal profile reached its maximumat Φ ~ 66° (GILL). The frequency of ~3.6 mHzbecame predominant when the substorm intensitybecame maximal (Fig. 4b). The maximal intensity wasobserved at the same geomagnetic latitude (66°). Timeshifts between the oscillations at different stations(ISLL–GILL–FCHU) (Fig. 3) indicate that the waveapparently propagated from low latitudes to higher ones.

According to the data of the GILL–PBO pair ofmagnetic stations, which are spaced by ΔΛ = 19.5°and are located at close geomagnetic latitudes (ΔΦ < 1°),the azimuthal wave number of observed Pc5 pulsationswas determined from formulas m = Δϕ/ΔΛ based onthe cross correlation function time shift (Δt) (hereΔϕ = Δt/T × 360°, and T is the oscillation perioddetermined from a spectral analysis). From 1230 to1320 UT, the correlation coefficient reached ~0.8 atΔt = 60 s, according to which the azimuthal wavenumber is m ≈ 2.7 for 5 mHz (T ~ 200 s) and m ≈ 4.0for 3.6 mHz (T ≈ 278 s) at 1230–1330 UT. Such an mvalue actually corresponds to the large�scale azi�muthal structure of geomagnetic oscillations, i.e., tothe toroidal mode propagating anti�sunward.

3.3. Pc5 geomagnetic pulsations in the magneto�sphere. Pulsation amplitudes observed in the mag�netosphere strongly depend on the satellite position ona field line relative to the oscillation longitudinal struc�ture. Pc5 pulsations are as a rule standing Alfvén wavesbetween conducting ionospheres. The fundamentaloscillation mode near the field line top, electric field( ) and compressional magnetic field ( ) compo�nents have anti�node, and a node of transverse mag�netic components ( ). On the contrary, the secondharmonic has a anti�node and a node of compo�nents and

On the whole, good agreement is observed betweenPc5 pulsations visible on the ground and in the mag�netosphere. The polarization structure of magneticfield oscillations registered on the GOES�8 geosta�tionary satellite (Fig. 5) corresponds to toroidal Alfvénoscillations: the azimuthal (Вn) component (~10 nT)predominates over the radial (Вe) component (~1 nT).In this case the oscillations manifest themselves(although insignificantly) in the magnetospheric mag�netic field magnitude (Bt), which may be related to theweak coupling between modes. The magnetosphericpulsations observed on GOES�8 are the combinationof the fundamental mode with f ~ 3.6 mHz with thesecond harmonic with f ~ 5.0 mHz (Fig. 6). In thetransverse magnetic components, the second har�monic predominates, and the fundamental harmonicis weaker and is mainly observed in the Be and Bt com�

E⊥ B�

.B⊥

B⊥

B� .E⊥

46


BELAKHOVSKY et al.

ponents. The relationship between the intensities ofthese harmonics is strongly masked by the satelliteposition relative to the wave nodes and anti�nodes;therefore, the second harmonic predominates duringthe entire event since the satellite is located near thewave anti�node. According to the ground�based obser�vations, the relationship between the harmonics variesduring the event: the second harmonic predominated atthe event onset, and the fundamental mode intensifiedafter the substorm onset and particle injection.

A close relationship is observed between geomag�netic pulsations in the azimuthal component (Bn) on

GOES�8 and pulsations in the geomagnetic field Xcomponent at PBO station (Fig. 5), which is locatednear the conjugate point with respect to GOES�8. Thecross correlation coefficient between these pulsationsis ~0.7 at a time delay of Δt ~ 0 (not shown). The coin�cidence of the oscillations in the geomagnetic fieldazimuthal component in the magnetosphere and inthe geomagnetic field horizontal component on theEarth’s surface corresponds to a rotation of an Alfvénoscillation polarization ellipse by π/2 when the wavecrossed the ionosphere. The oscillation frequencies onGOES�8 and at PBO conjugate station coincide (f ~5.0 mHz) at the event onset (1130–1230 UT) and dif�

0.96

16001100 1500140013001200

0.98

1.00

1.020

50010001500200025003000

0

0.5

1.0

1.5

2.0

2.5

1000200030004000500060007000

200

400

600

800

1000

January 1, 2000

AE

LANL – 1989, e

GILL, rio

GILL, X

GILL, 557.7, 65 deg

50–75 keV75–105 keV105–150 keV

Inte

nsi

ty,

RA

bso

rpti

on

, d

BF

lux

AE

, n

T

UT

Mag

net

ic f

ield

, n

T 1

04

Fig. 2. The geophysical conditions during the analyzed event of January 1, 2000 (1100–1600 UT): the AE index, the electronintensities in the energy channels (50–75 keV, 75–105 keV, and 105–150 keV) on LANL�89; riometer absorption, the auroralintensity in the 557.7 nm band and the geomagnetic field (the X component) at GILL.



fer during the second phase of the event (f ~ 3.6 and~5.0 mHz). We assume that this is related to thechange in the relationship between the intensities ofthe fundamental and second harmonics. Pc5 pulsa�tions are already observed insignificantly (not shown) onthe GOES�10 geostationary satellite (MLT = UT – 9),which was at a distance corresponding 4 h MLT fromGOES�8 (MLT = UT – 5).

3.4. Auroral Pc5 pulsations and their relation togeomagnetic pulsations and pulsations in riometerabsorption. Pc5 geomagnetic pulsations are accompa�nied by Pc5 pulsations in the oxygen 557.7 and 630.0 nmemissions, nitrogen 471.0 nm emission, and riometerabsorption (Figs. 7, 8). Observed Pc5 auroral pulsa�tions are periodic discrete auroral arcs. The modula�tion effects appeared on GILL from ~1220 UT, when

–30

1100 1130 1200 1230 1300 1330 1400

–20–10

01020

–100

–50

0

50

100–200

–100

0

100

200

–100–50

050

–150

100150

–200

0

100

–100

–200

050

–100–150

–50

100January 1, 2000

RANK, 72.22

ESKI, 70.52

FCHU, 68.32

GILL, 66.03

ISLL, 63.62

PINA, 59.98

UT

Fig. 3. Variations in the geomagnetic field X component according to the PINA–ISLL–GILL–FCHU–ESKI–RANK dataalong 330° longitude on January 1, 2000, 1100–1400 UT with a remote trend. The station geomagnetic latitude is shown in brackets.

48


BELAKHOVSKY et al.

the general background of auroral luminosity in 557.7nm, and riometer absorption increased (Fig. 7). Dis�tinct auroral pulsations were also observed at the high�latitude edge of the PINA photometer field of view (notshown). The instant when auroral luminosity periodicpulsations started intensifying corresponds to theinstant when the energetic electron cloud arrived atthe observation sector according to the LANL data.The intensity of auroras and pulsations was maximal inthe 557.7 nm emission. In the 486.0 nm proton emis�sion, the intensity of auroras was weak, and Pc5 pulsa�tions were no observed. At FSMI station, which islocated at a distance of ΔΛ ~ 26° from GILL atapproximately the same latitude, auroral Pc5 pulsa�tions are observed, but the correlation between pulsa�tions is weak. Therefore, it was impossible to deter�

mine the azimuthal propagation characteristics basedon photometric data.

Pc5 pulsations in different auroral emissions(557.7, 630.0, and 471.0 nm) have a close frequencycomposition and correlate well with one another. ForGILL, the correlation coefficient (r) between the vari�ation in the geomagnetic field X component andauroral intensity (557.7 nm) at 1230–1330 UT reaches0.72 at a time shift of Δt = 0 (not shown). A cross�spectral analysis for the same time interval indicatedthat a very high coherence exists (γ ~ 0.97) betweenvariations in the geomagnetic field X component andthe auroral intensity (557.7 nm) near the spectral max�imums (3.1 and 3.6 mHz). The phase difference nearthe spectral maximums varied insignificantly and was

0

61 5432

5.0 × 105

1.0 × 106

1.5 × 106

2.0 × 106

2.5 × 106

0

2.0 × 105

4.0 × 105

6.0 × 105

8.0 × 105

1.0 × 106

1.2 × 106

1.4 × 106 GILLFCHUESKIRANKISLLPINA

GILLFCHUESKIRANKISLLPINA

Frequency, mHz

Spe

ctra

l pow

erS

pect

ral p

ower

1120–1240 UT

1230–1330 UT

(a)

(b)

Fig. 4. Spectral analysis of geomagnetic Pc5 pulsations (the X component) registered on the N–S line (330°) at (a) 1120–1240 UT and (b) 1230–1320 UT.



Δφ ~ 20°; Pc5 pulsations in the magnetic field wereahead of auroral pulsations in this case. However, weshould bear in mind that the phase shift estimatesobtained from the spectral cross analysis correspondto the time shift smaller than the data discretizationinterval (1 min): Δ t = T(Δφ/360°) ~ 15 s at a variationperiod of T = 280 s.

A characteristic structure is observed for auroralPc5 pulsations in the 557.7 and 630.0 nm emissions(Figs. 7, 8), where each brightness element is inclinedto the right. This corresponds to the arrival of particles(which causes luminosity brightening), first to lowerlatitudes and then to higher ones. The apparent prop�agation from low latitudes to higher ones reflects the

phase structure of field�aligned currents that are trans�ported by a wave in the Alfvén resonance region. Inaddition, spectral analysis indicates that the frequencyof auroral Pc5 pulsations decreases with increasinggeomagnetic latitude. Thus, at 1230–1330 UT, thefrequency of auroral Pc5 pulsations decreases from 3.6to 3.1 mHz at geomagnetic latitudes of 63.5° to 69.0°(the 557.7 nm emission) and 60.7° to 71.1° (the 630.0 nmemission). The spatial structure of auroral Pc5 pulsa�tions (the propagation from low latitudes to higherones and a decrease in the frequency with increasinggeomagnetic latitude) is similar to the structure ofgeomagnetic Pc5 pulsations.

108

13301100 1300123012001130

110

112

114

116

1.08

1.09

1.10

1.11

–8

–6

–4

–2

0

2

40

42

44

46

48

50

Mag

net

ic fi

eld,

nT

× 1

04

Mag

net

ic fi

eld,

nT

Mag

net

ic fi

eld,

nT

Mag

net

ic fi

eld,

nT

GOES – 8, Be

GOES – 8, Bn

PBQ, X

GOES – 8, Bt

UT

January 1, 2000

Fig. 5. Variations in the magnetospheric magnetic field according to the GOES geostationary satellite data for January 1, 2000(the Вe and Вn components and the field magnitude Вt) and the geomagnetic field X component at PBQ magnetically conjugatestation.

50


BELAKHOVSKY et al.

Detailed analysis indicates that far from all trains ofgeomagnetic Pc5 pulsations are accompanied bysimultaneous auroral pulsations. This fact indicatesthat geomagnetic Pc5 pulsations cause correspondingpulsations in auroras rather than the reverse, as wasassumed by Yamamoto et al. (1988), due to a localchange in the ionospheric conductivity by pulsatingparticle fluxes.

The relationship between the intensities of differ�ent emissions makes it possible to estimate the energyof precipitating electrons. Oxygen atoms in the 557.7 nmemission are more excited when higher�energy electronsprecipitate into the ionosphere than in the case whensuch atoms in the 630.0 nm emission are excited. For thebackground luminosity, the relationship between theintensities of these emissions is I557.7/I630.0 ~ 2–3 (Fig. 9).When auroral activity (accompanied by luminositypulsations) intensifies, the intensity ratio (I557.7/I630.0)

sharply increases to 12–16, although the intensities ofboth auroral emissions increase (Fig. 9). Consequently,geomagnetic Pc5 pulsations cause not only periodicmodulation of the precipitating particles but also resultin an additional periodic electron acceleration.

Auroral and geomagnetic Pc5 pulsations areaccompanied by simultaneous pulsations in the riom�eter absorption at GILL (Fig. 7). Pc5 pulsations in theriometer absorption were also observed at all of theother stations (FSMI and MCMU) spaced along thelongitude. In this case far from all Pc5 pulsations inabsorption are accompanied by pulsations in auroras.

Riometer oscillations look like short�term bursts atindividual instants. Spectral analysis indicates that thespectral maximums in the variations in the magneticfield, riometer absorption, and auroral intensity coin�cide at GILL (Fig. 10). Spectral cross analysis of Pc5pulsations in auroras (557.7 nm) and absorption at

5.0 × 105

61 5432

1.0 × 106

1.5 × 106

500

1000

1500

2000

2500

Spe

ctra

l pow

er

Spe

ctra

l pow

er

GOES – 8

PBQ

X

Z

Y

Bn

Be

Bt

January 1, 2000, 1230–1330 UT

Frequency, mHz

Fig. 6. Spectral analysis of geomagnetic Pc5 pulsations registered on GOES�8 (the Вe and Вn components and the field magnitude Вt)and at PBQ magnetically conjugate station (the X, Y, and Z components) at 1230–1330 UT.



1230–1330 UT indicates that coherence is γ ~ 0.6 atf ~ 3.6 mHz. In this case pulsations in the absorptionare ahead of those in auroras: the phase delay is Δφ ~ 50°,and r ~ 0.65 at Δt ~ 1 min. On the whole, stable phaserelationships between Pc5 pulsations in auroras andabsorption are not observed, and the correlationbetween these pulsations is low. This may indicate thatthe mechanisms by which the riometer absorption andauroral intensity are modulated by Pc5 pulsations aredifferent.

The modulation depth, which is the ratio of theoscillation amplitude to the total signal level, is an

important parameter of oscillation processes. For theanalyzed event, the modulation depth for the geomag�netic field is ΔB/B0 ≈ 0.5% at В0 = 56000 nT at 100 kmaltitude (according to the IGRF model) and ΔВ ~ 300 nT.For auroral pulsations, the modulation depth isΔI557.7/I557.7 ≈ 80% in the 557.7 nm emission,ΔI630.0/I630.0 ≈ 30% in the 630.0 nm emission, andΔI471/I471 ≈ 77% in the 471.0 nm emission; ΔR/R ≈50% for riometer absorption. Thus, the modulationdepth for Pc5 oscillations in the fluxes of precipitatingelectrons markedly exceeds the modulation depth inthe geomagnetic field, even though geomagnetic Pc5

013301130 130012301200

0.5

1.0

1.5

2.0

2.5

9.60 × 103

9.70 × 103

9.80 × 103

9.90 × 103

1.00 × 104

1.01 × 104

500

1000

1500

2000

2500

3000

64

66

68

70

Abs

orpt

ion

, dB

Inte

nsi

ty, R

Geo

mag

net

ic la

titu

de, d

eg

January 1, 2000

GILL, X

GILL, rio

557.7, 65 deg

UT

Fig. 7. Keogram of the auroral intensity according to the GILL photometer data in the 557.7 nm emission; the variations in theauroral intensity at Φ = 65°; the variation in the geomagnetic field X component and riometer absorption at GILL at 1130–1330 UT.

Mag

net

ic fi

eld,

nT

52


BELAKHOVSKY et al.

50

13301100 1300123012001130

100

150

200

200

300

400

600

800

66

68

70

64

66

68

70

72G

eom

agn

etic

lati

tude

, deg

Geo

mag

net

ic la

titu

de, d

egIn

ten

sity

, R

Inte

nsi

ty,

R

January 1, 2000

630.0 nm, 69.4 deg

471 nm, 68.01 deg

UT

Fig. 8. Keogram of the auroral intensity according to the GILL photometer data in the 630.0 and 471 nm emissions; the variationsin the auroral intensity at Φ = 68°.

oscillations cause simultaneous oscillations in aurorasand riometer absorption.

The latitudinal distribution of the variation spectralamplitude at a specified frequency makes it possible todetermine the latitudinal maximum half�width. For1230–1330 UT, geomagnetic pulsations at a fre�quency of 3.6 mHz have a maximum at Φ ~ 66° lati�tude with a half�width of ΔΦ ~ 3° (Fig. 11). The latitu�

dinal distribution of the spectral power at the same fre�quency of the 630.0 nm auroral emission has amaximum at Φ ~ 69.5° latitude with the ΔΦ ~ 1.5°half�width; such a distribution of the 557.7 nm emis�sion has a maximum at Φ ~ 68° with a half�width ofΔΦ ~ 1.2° (Fig. 11). Thus, the power maximum is at68°(69.5°) and 66° geomagnetic latitude for auroraland geomagnetic Pc5 pulsations, respectively. In this



case the half�width of the latitudinal distribution forauroral pulsations is smaller than the half�width forgeomagnetic pulsations by 1°.

4. MODULATION OF THE REGIONOF AURORAL ELECTRON ACCELERATION

BY ALFVÉN WAVE

Researchers often relate the periodic accelerationand precipitation of auroral electrons to the longitudi�nal electric field E|| of a dispersion Alfvén wave. For theobserved transverse scale of the auroral luminosityregion (>1°), the dispersion corrections, which are

related to a finite Larmor radius or to an inertial elec�tron length, are too small for the appearance of pro�nounced E|| in an Alfvén wave. The appearance of E||

may be related to the characteristic feature of theauroral oval: to the existence of the auroral accelera�tion region (AAR), where an electric potential dropcaused by mirror resistance along a field line is formedat ~1–2 RE altitudes. This nonresistive potential dif�ference along field lines makes it possible to maintainthe current flowing from the ionosphere, even thoughthe cone angles of auroral field lines are small. Thiscurrent ( ) and the potential difference (ΔΦ) arerelated by the nonlocal ohmic relation (the Knight

||j

213301100 1300123012001130

4

6

8

10

12

14

16100

150

200

250

300

1000

2000

3000In

ten

sity

, R

Inte

nsi

ty,

R

January 1, 2000 GILL. Photometer CGL = 65.53 deg

557.7 nm

630.0 nm

557.7/630.0

UT

Fig. 9. The auroral intensity in the 557.7 and 630.0 nm emissions at GILL and the 557.7/630.0 intensity ratio for 1100–1330 UT.

54


BELAKHOVSKY et al.

relation) where Q is a field line mirrorresistance (Vogt, 2002). Typical values of a longitudi�nal mirror resistance are Q = 107–109 Ω m2, for whichthe potential difference should be ΔΦ ~ 0.01–10 keVin order to maintain auroral currents with a density of

~ 10–6–10–5 A m–2. In particular, the relationship

( ) between the average energy of precipitatingelectrons W0 (measured with respect to auroral emis�sions) and the total energy (Wt) transported by auroralelectrons points to the existence of AAR (Ono andMorishima, 1994). Such a “quasi�ohmic” relation istypical of electrons accelerated in AAR.

It was theoretically shown that the interactionbetween an Alfvén wave and AAR results in theappearance of a pulsating longitudinal electric fieldthat causes an additional acceleration of auroral elec�trons (Fedorov et al., 2004). Magnetospheric Alfvénwaves penetrating AAR cause periodic modulation ofthe potential difference (ΔΦ). In the scope of a thin

| |,Q jΔΦ =

| |j2

0~tW W

AAR�layer approximation, the amplitude of theseoscillations, normalized to the magnetic field (В) of anAlfvén wave incident on AAR, is defined as

(1)

where = is the

Alfvén wave conductivity, is the Alfvénresistive scale, and VA is the Alfvén velocity. Relation�ship (1) indicates that the modulation of the potentialdifference by an Alfvén wave strongly depends on thewave transverse scale (this dependence is contained in

function). For ULF disturbances in a sunlitionosphere ( ), function reaches themaximal value for transverse scales Thus,the extreme values are Scale λA at typicalQ for ΣA = 1 Ω–1 (VA = 800 km/s) can vary from 3 to30 km. The additional energy (~eΔΦ), which is sup�

A( , ),K T kB

Φ ⊥

ΔΦ= λ ω

−

Φ = μ ΣA1 1 2( )K Q A A,Vλ A A

1( )V −

Σ = μ

A AQλ = Σ

A( )T k⊥λ

APΣ Σ� A( )T k⊥λ

A ~ 1.k⊥λ

.B KΦΔΦ =

10

1 2 3 4 5 6

20

30

40

2.0 × 105

4.0 × 105

6.0 × 105

8.0 × 105

1.0 × 106

1.2 × 106

1.4 × 106

5.0 × 105

1.0 × 106

1.5 × 106

2.0 × 106

2.5 × 106

January 1, 2000, 1230–1330 UT

GILL, X

GILL, Y

GILL, I5577/100

GILL, I6300

GILL, I471

GILL, rio

Frequency, mHz

Spe

ctra

l pow

erS

pect

ral p

ower

Spe

ctra

l pow

er

Fig. 10. Spectra of pulsations for 1230–1330 UT at GILL, the geomagnetic field (the X and Y components), the intensity of the630.0 nm, 557.7 nm (reduced by a factor of 100), and 471.0 nm emissions auroral emissions and riometer absorption.



plied to auroral electrons by the Alfvén wave field withan amplitude of B = 40 nT, reaches ~1–10 keV at opti�mal transverse scales (ТΦ = 1). Thus, the additionalmodulated acceleration in AAR can cause auroralluminosity brightening, which should manifest itselfduring photometric observations but cannot causeacceleration and precipitation of energetic electronswith E > 30 keV, which are responsible for riometerabsorption variations. The described electron acceler�ation mechanism can operate only in the auroral oval,where AAR is formed.

The auroral modulation peculiarity consists in thefact that auroral luminosity modulation by geomag�netic Pc5 pulsations is a relatively rare phenomenon,though these pulsations are often observed at aurorallatitudes. According to this mechanism, this circum�stance is related to the fact that the Alfvén wave trans�

verse scale ( ) should be optimally combined withthe AAR properties characterized by parameter λA. If

(e.g., λA = 10 km, and 100 km), the

1~ k −

⊥

A1k −

⊥ λ�1k −

⊥ =

energy that is supplied by geomagnetic Pc5 pulsationsto auroral electrons is low, although such pulsationsshould be clearly defined on the Earth’s surface. If the

Alfvén wave scale is optimal but small, (e.g.,

λA = 10 km, and 10 km), the ground magneticresponse of such waves will be insignificant due tostrong ionospheric shielding, since this scale is smallerthan the ionospheric altitude (h ~ 100 km). At thesame time, such waves will effectively periodicallymodulate auroral electron acceleration. Ground andauroral Pc5 pulsation will be observed simultaneouslyonly at favorable parameters of AAR and waves when

.

5. DISCUSSION

In the analyzed event, the spatial–time structure ofauroral Pc5 pulsations in different emissions reflectsthe field resonance structure of simultaneouslyobserved geomagnetic Pc5 pulsations: local intensifi�

A1k −

⊥ λ�

1k −

⊥ =

A1 ,k h−

⊥ λ ≥�

5

7260 62 64 66 68 70

10

15

20

25

5

7260 62 64 66 68 70

10

15

20

25

30

0

150

7260 62 64 66 68 70

100

200

250

300S

pect

ral a

mpl

itud

eS

pect

ral a

mpl

itud

eS

pect

ral a

mpl

itud

e

Geomagnetic latitude, deg

January 1, 2000, 1230–1330 UT. 3.61 mHz

557.7

630.0

X

Fig. 11. Latitudinal distribution of the variation spectral amplitude at a frequency of 3.6 mHz for 1230–1330 UT: Pc5 pulsationsin the 557.7 nm auroral emission, Pc5 pulsations in the 630.00 nm auroral emission, and geomagnetic Pc5 pulsations.

56


BELAKHOVSKY et al.

cation in the same latitudinal region and propagationfrom low latitudes to higher ones. The latitudinalamplitude–phase structure of geomagnetic Pc5 pulsa�tions indicate that these pulsations are resonance:local Alfvén oscillations of a field line are excited by anexternal harmonic source. The Kelvin–Helmholtzinstability at the magnetopause, which develops whenthe solar wind velocity is high and is followed by thepropagation of a fast magnetosonic wave deep into themagnetosphere and field line resonance at individualL shells, is possibly responsible for the generation ofobserved geomagnetic Pc5 pulsations. The observedgeomagnetic Pc5 pulsations actually have toroidalpolarization in the magnetosphere and are character�ized by small azimuthal wave numbers (m ≈ 2.7).When an Alfvén wave propagates trough the iono�sphere, a polarization ellipse should turn by 90°; i.e.,the field azimuthal component (Bn), which is mostpronounced in resonance oscillations in the magneto�sphere, is observed as a meridional (N–S) X compo�nent. In the resonance region, the N–S profile of theX component amplitude sharply locally increases, andthe phase reaches ~π when crossing the latitudinalmaximum. In the same region, the vertical magneticcomponent (Z) has a local spatial maximum (Pil�ipenko et al., 2000). The ULF wave field structurepredicted by a resonance theory agrees well with thesynchronous observations of the magnetic field andauroral luminosity Pc5 pulsations on the N–S line ofstations (Roldugin and Roldugin, 2008) and is con�firmed by our observations.

Auroral Pc5 pulsations are observed in a narrowerlatitudinal interval than geomagnetic pulsations. Thelatitudinal scale of the auroral electron accelerationregion apparently corresponds to the characteristicdimension of the resonance region in the ionosphere(Δxi). The resonance structure half�width on theEarth’s surface (Δx) should be Δx = Δxi + h, where h =120 km is the altitude of the ionospheric conductinglayer (Pilipenko et al., 2000). This relationship agreeswith a difference (~1°) in the measured half�widths ofthe latitudinal maximums of magnetic and auroralpulsations.

In addition to the modulation of auroras, whichdevelop during auroral substorms, the auroral Pc5 pul�sations considered in this work are a simple and reli�able marker of the region where substorms originate(Rae et al., 2014). The existence of resonance oscilla�tions with f ~ 3.1–3.6 mHz in the auroral luminosityintensification region unambiguously indicates thatthis auroral substorm develops in the region of closedfield lines.

Several works proposed an inverse mechanism ofthe interrelation between auroral and magnetic pulsa�tions: the excitation of geomagnetic Pc5 pulsation dueto a change in the local ionospheric conductivity dur�ing modulated energetic particle precipitation (Yama�moto, 1988). Our observations and those of others

show that far from all trains of geomagnetic Pc5 pulsa�tions are accompanied by auroral Pc5 pulsations. Weassume that this indicates that geomagnetic pulsationsare primary as compared to auroral ones. In addition,if any ionospheric source actually excites geomagneticPc5 pulsations, these pulsations (as observed byground magnetometers) should represent the super�position of pulsations caused by magnetospheric andionospheric sources. According to this viewpoint, aconsiderable difference should exist between Pc5 pul�sations observed in the magnetosphere and on theground. However, for the considered case, good agree�ment exists between Pc5 pulsations observed byground magnetometers and magnetometers in geosta�tionary orbit.

Stable phase relations are not observed betweenpulsations in riometer absorption and auroras, whichmay indicate that energetic (~10–100 keV) andsuprathermal (~0.1–1 keV) electrons precipitate dif�ferently. The field of magnetospheric Pc5 oscillationsis apparently insufficient for the acceleration of elec�trons to energies that can change riometer absorption(>30 keV). Therefore, Pc5 pulsations were generateddifferently in riometer absorption and auroras. Themodulated precipitation of energetic electrons, whichmanifests itself as pulsations in riometer absorption,can be caused by compression of the magnetosphericmagnetic field (b||) in the wave field (Coroniti andKennel, 1970). This disturbance of the adiabaticmotion of energetic electrons and their precipitationinto the loss cone may be related to scattering by theVLF noise generated by injected electrons. A short�term burst of riometer oscillations can occur duringthe synchronization of two oscillatory systems: Alfvénwaves in the magnetospheric resonator and the ener�getic electrons + VLF noise system (Belakhovsky andPilipenko, 2010).

Various models based on the concepts of oscilla�tions in the magnetospheric Alfvén resonator wereproposed in order to interpret the auroral arcs thatperiodically move poleward. Lyatsky et al. (1999) pro�posed a qualitative explanation based on the effect of“phase mixing” between field lines oscillating withdifferent frequencies. However, this model is based onextremely simplified views of the resonance regionspatial structure as a set of isolated noninteractingmagnetic shells and cannot be accepted. The interpre�tation relating the poleward arc motion to the phasepropagation of Pc5 waves across magnetic shells in theAlfvén resonance region seems to be most probable(Samson et al., 1992; Xu et al., 1993). Geomagneticpulsations should propagate poleward in the regionswhere the Alfvén period TA(L) increases with increas�ing L shell. The structure of pulsating field�alignedcurrents, which are transported by an Alfvén wave inthe resonance region, is such that the outflowing cur�rent (transported by electrons precipitating into theionosphere) first originates at the resonance region



equatorward boundary, subsequently increases andmoves toward the region center, and then reaches thepoleward boundary of the resonance region and weak�ens. Therefore, two arcs can sometimes be observed atdifferent edges of the resonance region.

Periodic acceleration and precipitation of auroralelectrons can supposedly be caused by a longitudinalelectric wave field (E||). Such a field can originate in anAlfvén wave due to small�scale kinetic or dispersioneffects (Stasiewicz et al., 2000); however, such small�scale waves would be completely shielded by the iono�sphere from ground magnetometers. The kinetic andnonlinear effects, which result in the appearance of E||

and the following acceleration of suprathermal elec�trons along field lines, can develop in the localizedfield�aligned currents in the Alfvén resonance region(Samson, 2003). This mechanism is formally indepen�dent of the auroral oval and can operate in any mag�netospheric region where the frequency of field lineAlfvén oscillations coincides with the external sourcefrequency.

An abrupt increase in the oxygen emission intensityratio (557.7/630.0) during the appearance of auroralPc5 pulsations indicates that suprathermal electronsperiodically accelerate in the pulsation field. Thisacceleration may be related to the proposed modula�tion of the region where the potential drops alongauroral field lines (AAR) by Alfvén waves. Theabsence of a pronounced delay between field varia�tions and the particle fluxes registered by a photometer(not more than 20 s at a 1�min data discretization)may indicate that auroral electrons supposedly accel�erate at comparatively low altitudes (~1RE) above theionosphere rather than at the field line top. Indeed,geomagnetic pulsations are standing Alfvén oscilla�tions along a field line; therefore, wave fields on a sat�ellite and on the Earth’s surface should be in�phase, ifdissipation is ignored. If particle modulation/acceler�ation takes place near the equatorial plane, luminositywill lay behind field variations by the particle traveltime to the lower ionosphere.

It is clear that the considered scenario does notinclude all possible mechanisms by which auroralelectrons are modulated and accelerated by Pc5 pulsa�tions. Thus, Saka et al. (2014) detected pulsatingbursts of auroral luminosity synchronously with pow�erful geomagnetic Pc5 pulsations in the dawn sector,which are observed after an increase in electron andproton fluxes in geostationary orbit. These researchersassumed that these pulsating auroras are the manifes�tation of a poloidal Alfvén wave with a considerablemagnetic field compression component and that themodulation itself is caused by period variations in themagnetic field value in the magnetospheric equatorialplane.

6. CONCLUSIONS

The auroral periodic pulsations observed in the557.7, 630.0, and 471.0 nm emissions result not onlyfrom electron flux modulation by magnetospheric tor�oidal Alfvén waves in the Pc5 range but also from theperiodic wave acceleration of electrons above theauroral oval. Auroral pulsations reflect the resonancestructure of toroidal geomagnetic Pc5 pulsations (anincrease in the period with increasing geomagnetic lat�itude and a phase propagation from low latitudes tohigher ones). In this case geomagnetic Pc5 pulsationsare primary, as compared to auroral Pc5 pulsations.

Although the frequencies of Pc5 pulsations in auro�ras and riometer absorption are close to each other, thecorrelation between them is low, which may indicatethat electrons of different energies have the samesource but are modulated differently. We assumed thatthe interaction between an Alfvén wave and the poten�tial difference region above the auroral oval (AAR)results in the appearance of a pulsating longitudinalelectric field. This longitudinal field causes an addi�tional acceleration of auroral electrons, which cancause pulsations of auroral luminosity in the 630.0 and557.7 nm emissions. The energy is transferred from anAlfvén wave to auroral electrons most effectively whenthe wave transverse scales are about the Alfvén dissipa�tive length (λA), which is several dozen kilometers fortypical values of the field line mirror resistance. How�ever, the longitudinal field in AAR is insufficient foracceleration and precipitation of higher�energy elec�trons, which are responsible for riometer absorptionvariations, and this field is modulated differently.

ACKNOWLEDGMENTS

We are grateful to members of the NORSTAR,CARISMA, and GOES centers for the presented dataof the corresponding projects and to Geoff Reeves andhttp://www.lanl�epdata.lanl.gov for the presentedLANL satellite data.

This work was supported by a president of the Rus�sian Federation, grant MK�4210.2015.5 (BV); RussianFoundation for Basic Research, project no. 15�45�05108 (SS); Presidium of the Russian Academy of Sci�ences, program 9 (PV); and Research Council of Nor�way, grant 246725 (LD).

REFERENCES

Baishev, D.G., Barkova, E.S., and Yumoto, K., Opticalobservations of large�scale undulations in the 23rd cycleof solar activity, Geomagn. Aeron. (Engl. Transl.), 2012,vol. 52, no. 2, pp. 197–203.

Belakhovsky, V.B. and Pilipenko, V.A., Excitation of Pc5pulsations of the geomagnetic field and riometricabsorption, Cosmic Res., 2010, vol. 48, no. 4, pp. 319–334.

58


BELAKHOVSKY et al.

Coroniti, F.V. and Kennel, C.F., Electron precipitationpulsations, J. Geophys. Res., 1970, vol. 75, pp. 1279–1289.

Fedorov, E.N., Pilipenko, V.A., Engebretson, M.J., andRosenberg, T.J., Alfvén wave modulation of the auroralacceleration region, Earth, Planets Space, 2004, vol. 56,pp. 649–661.

Keiling, A., Wygant, J.R., Cattell, C., Peria, W., Parks, G.,Temerin, M., Mozer, F.S., Russell, C.T., andKletzing, C.A., Correlation of Alfvén wave Poyntingflux in the plasma sheet at 4–7 RE with ionosphericelectron energy flux, J. Geophys. Res., 2002, vol. 107,no. A7, pp. SMP24�1–SMP24�13.

Liou, K. and Takahashi, K., Observations of field line reso�nance with global auroral images, J. Atmos. Sol–Terr.Phys., 2013, vol. 105, pp. 152–159.

Lyatsky, W., Elphinstone, R.D., Pao, Q., and Cogger, L.L.,Field line resonance interference model for multipleauroral arc generation, J. Geophys. Res., 1999, vol. 104,no. A1, pp. 263–268.

Milan, S.E., Yeoman, T.K., Lester, M., Moen, J., andSandholt, P.E., Post�noon two�minute period pulsat�ing aurora and their relationship to the dayside convec�tion pattern, Ann. Geophysicae, 1999, vol. 17, pp. 877–891.

Oieroset, M., Sandholt, P.E., Luehr, H., Denig, W.F., andMoretto, T., Auroral and geomagnetic events atcusp/mantle latitudes in the prenoon sector during pos�itive IMF by conditions: Signatures of pulsed magneto�pause reconnection, J. Geophys. Res., 1997, vol. 102,pp. 7191–7205.

Olson, J.V., Rostoker, G., and Olchowy, G., A study ofconcurrent riometer and magnetometer variations inthe Pc4–5 pulsation band, J. Geophys. Res., 1980, vol. 85,pp. 1695–1702.

Ono, T. and Morishima, K., Energy parameters of precipi�tating auroral electrons obtained by using photometricobservations, Geophys. Res. Lett., 1994, vol. 21, no. 4,pp. 261–264.

Pilipenko, V., Vellante, M., and Fedorov, E., Distortion ofthe ULF wave spatial structure upon transmissionthrough the ionosphere, J. Geophys. Res., 2000, vol. 105,pp. 21225–21236.

Pilipenko, V., Belakhovsky, V., Kozlovsky, A., Fedorov, E.,and Kauristie, K., ULF wave modulation of the iono�spheric parameters: Radar and magnetometer observa�tions, J. Atmos. Sol.–Terr. Phys., 2014, vol. 108, pp. 68–76.

Rae, I.J., Mann, I.R., Dent, Z.C., Milling, D.K., Dono�van, E.F., and Spanswick, E., Multiple field line reso�nances: Optical, magnetic and absorption signatures,Planet. Space Sci., 2007, vol. 55, pp. 701–713.

Rae, I.J., Murphy, K.R., Watt, C.E.J., Rostoker, G.,Rankin, R., Mann, I.R., Hodgson, C.R., Frey, H.U.,Degeling, A.W., and Forsyth, C., Field line resonancesas a trigger and a tracer for substorm onset, J. Geophys.Res., 2014, vol. 119, pp. 5343–5363.

Roldugin, V.C. and Roldugin, A.V., Pc5 pulsations on theground, in the magnetosphere, and in the electron pre�cipitation: Event of 19 January 2005, J. Geophys. Res.,2008, vol. 113, no. A4, p. A04222.

Saka, O., Hayashi, K., Klimushkin, D.Yu., and Mager, P.N.,Modulation of auroras by Pc5 pulsations in the dawnsector in association with reappearance of energeticparticles at geosynchronous orbit, J. Atmos. Sol.–Terr.Phys., 2014, vols. 110–111, pp. 1–8.

Samson, J.C., Wallis, D.D., Hughes, T.J., Creutzberg, F.,Ruohontemi, J.M., and Greenwald, R.A., Substormintensifications and field line resonances in the night�side magnetosphere, J. Geophys. Res., 1992, vol. 97,pp. 8495–8518.

Samson, J.C., Rankin, R., and Tikhonchuk, V.T., Opticalsignatures of auroral arcs produced by field line reso�nances: Comparison with satellite observations andmodeling, Ann. Geophysicae, 2003, vol. 21, pp. 933–945.

Sarris, T.E., Loto’aniu, T.M., Li, X., and Singer, H.J.,Observations at geosynchronous orbit of a persistentPc5 geomagnetic pulsation and energetic electron fluxmodulations, Ann. Geophysicae, 2007, vol. 25, pp. 1653–1667.

Stasiewicz, K., Bellan, P., Chaston, C., Kletzing, C.,Lysak, R., Maggs, J., Pokhotelov, O., Seyler, C.,Shukla, P., Stenflo, L., Streltsov, A., and Wahlund, J.�E.,Small scale Alfvénic structure in the aurora, Space Sci.Rev., 2000, vol. 93, pp. 423–533.

Vogt, J., Alfvén wave coupling in the auroral current circuit,Surv. Geophys., 2002, vol. 23, pp. 335–377.

Vorobjev, V.G., Belakhovsky, V.B., Yagodkina, O.I.,Roldugin, V.K., and Hairston, M.R., Features of morn�ing time auroras during SC, Geomagn. Aeron. (Engl.Transl.), 2008, vol. 48, no. 2, pp. 154–164.

Xu, B.�L., Samson, J.C., Liu, W.W., Creutzberg, F., andHughes, T.J., Observations of optical aurora modu�lated by resonant Alfvén waves, J. Geophys. Res., 1993,vol. 98, no. A7, pp. 11531–11541.

Yamamoto, T., Hayashi, K., Kokubun, S., Oguti, T., andOgawa, T., Auroral activities and long�period geomag�netic pulsations: Pc5 pulsations and concurrent aurorasin the dawn sector, J. Geomagn. Geoelectr., 1988, vol. 40,pp. 553–569.

Translated by Yu. Safronov

Features of Pc5 Pulsations in the Geomagnetic Field ... · GEOMAGNETISM AND AERONOMY Vol. 56 No. 1...

Documents

Transcript of Features of Pc5 Pulsations in the Geomagnetic Field ... · GEOMAGNETISM AND AERONOMY Vol. 56 No. 1...