Magnetospheric Electrodynamics: Energetic Particle Signatures Of

Home , Geomagnetic storm, Magnetic cloud, Plasma sheet

MAGNETOSPHERIC ELECTRODYNAMICS: ENERGETIC PARTICLE SIGNATURES OF

GEOMAGNETIC STORMS AND SUBSTORMS

Tuija I Pulkkinen

Finnish Meteorological Institute, POBox 503, FIN-00101 Helsinki, Finland

ABSTRACT electrons and ions reside in the ring current and in

the van Allen radiation b elts.

This pap er reviews the present understanding of the

dynamics of two key magnetospheric disturbances, Substorms, b e- geomagnetic substorms and storms.

INTERBALL

ing the basic dynamic resp onse of the magnetosphere

to the varying solar wind conditions, are an imp or-

tant factor in our understanding of the solar wind

magnetosphere coupling pro cesses. On the other

{ POLAR

hand, the geomagnetic storms represent times when

the magnetosphere is under extremely disturb ed con-

ditions and are hence of key imp ortance for space

Solar wind Plasma sheet

weather. The present and future capabilities to fore-

SAMPEX

wintense events o ccur based cast when, where, and ho Geostationary s/c

L-shell

on solar or upstream solar wind observations are brie y addressed. GEOTAIL SOHO

WIND

words: Magnetosphere, storms, substorms, en- Key ACE

IMP-8 ergetic particles, coronal mass ejections.

Outer radiation belt

Ring current

1. INTRODUCTION

key issues for space weather can be set in the

The

form of two questions: 1 Event or no event; and

2 If an eventdoestake place, when do es it o ccur,

where do es it o ccur, and how intense is it. While

the rst question can be largely addressed by solar

Figure 1. Schematic of the near-Earth magnetosphere.

and/or solar wind monitoring, our capabilitytoan-

The tail lobes, the plasma sheet light shading and ring

swer the second question dep ends critically on our

current/radiation belts darker shading are shown. Typ-

understanding of the physical pro cesses b ehind the

ical locations of several of the ISTP satel lites are also

dynamic changes in the near-Earth space.

indicated.

The magnetosphere is the region of near-Earth space,

The ongoing International Solar Terrestrial Physics

where the dynamics is governed by the internal geo-

ISTP program provides an unprecedented opp ortu-

magnetic eld see Figure 1. The solar wind ow

nity to monitor the near-Earth environment Acuna ~

past the Earth distorts the dip ole eld to compress it

et al. 1995. Several spacecraft such as SOHO,

on the dayside and elongate it to a long geomagnetic

WIND, IMP-8, and ACE monitor the Sun and the

tail on the nightside. The geomagnetic tail plays a

solar wind almost continuously. Simultaneously, sev-

key role in magnetospheric dynamics; for example, it

eral spacecraft such as POLAR, INTERBALL, and

acts as an energy reservoir for the dynamic pro cesses.

GEOTAIL measure dynamic variations within the

The magnetosphere comprises distinct regions, which

magnetosphere. Furthermore, several op erational

all have their characteristic plasma prop erties: The

satellites at geostationary orbit or at low-altitude p o-

tail lob es at high latitudes are regions of low plasma

lar orbit together with a multitude of ground-based

density and energy, whereas the plasma sheet is char-

facilities provide key information of the magneto-

acterized by denser and hotter keV plasma. The

spheric and ionospheric variability. Hence, the up-

most hazardous region for technological systems is

coming solar maximum will be monitored with an

the inner magnetosphere, where trapp ed p opulations

impressive set of observational facilities that serveas

of high-energy from hundreds of keV to multi-MeV

well scienti c as op erational space weather purp oses.

The dynamic resp onse of the magnetosphere to vary- netic eld and strong enhancement of the uxes of en-

ing solar wind and interplanetary magnetic eld con- ergetic tens to hundreds of keV ions ?? and high-

ditions is the magnetospheric substorm Rostoker energy up to several MeV electrons in the outer

et al. 1980. Energy input from the solar wind van Allen radiation b elt Baker et al. 1998a. Such

is largely controlled by the interplanetary magnetic activity is clearly a key factor in the problems exp e-

eld orientation: During p erio ds of southward inter- rienced by space-b orne technological systems during

planetary eld, the energy input is enhanced and the disturb ed p erio ds Baker et al. 1996b.

energy extracted from the solar wind is stored in the

This pap er reviews present understanding of the dy-

magnetosphere in the form of magnetic eld energy

namics of the geomagnetic substorms and storms by

in the magnetotail. This is the substorm growth

presenting observations and interpretations of two

phase. After typically 30{60 min, the magnetotail

events, a substorm on Dec 10, 1996 and a storm on

undergo es a change of state from stable to unsta-

May 15, 1997. Both events have b een analyzed in

ble, and the stored energy is dissipated via a highly

detail elsewhere see Pulkkinen et al. 1998a and

dynamic pro cess. This substorm expansion phase in-

Pulkkinen et al. 1998c for the substorm event and

volves an injection of energetic tens to hundreds of

Baker et al. 1998b for the storm event, and only a

keV electrons and ions to the vicinity of the geosta-

short summary is presented here. Finally, the present

tionary orbit, strong electric currents in the auroral

and future capabilities to address question 2 are

regions, and rapid uctuations and con gurational

discussed.

changes of the magnetospheric magnetic eld. All

these phenomena are p otential space weather e ects.

The substorm pro cess ends when the energy dissipa-

tion ceases and the magnetosphere recovers its initial

2. GEOMAGNETIC SUBSTORMS: Dec 10, 1996

state after ab out two to four hours from the b egin-

ning of the event see Figure 2; for recent reviews see

McPherron 1991 and Baker et al. 1996a.

The substorm activity on Dec 10, 1996, 06{10 UT,

was driven by a p erio d of southward interplanetary

magnetic eld IMF. The magnetic eld measure-

ts from WIND showed a southward turning at men INTERPLANETARY MAGNETOPAUSE

FIELD

ab out 0637 UT see Figure 3. The top panel in Fig- PLASMA MANTLE

TAIL LOBE

ure 3 show the interplanetary magnetic eld IMF TO SUN PLASMA SHEET

60 Re

B comp onent. The following panels show the =

7 2 2

10 vB l sin =2 parameter in SI units Perreault

0 R

EARTH 100 Re DISTANT NEUTRAL LINE

&Akasofu 1978 and its time integral dt. The

GROWTH PHASE: ENERGY LOADING

parameter is a measure of the energy input from the

solar wind into the magnetosphere; a clear increase

is visible after the IMF southward turning. In the

plot, the parameter has b een time-shifted 7 min to

t for the travel time from the s/c lo cation to THIN CURRENT SHEET accoun

EXPANSION ONSET the subsolar magnetopause.

The following panel shows the AL index as deduced

from the CANOPUS ground magnetometer data. The electro jet indices are created from disturbances

NEAR-EARTH NEUTRAL LINE

in the North-South comp onent of auroral-zone mag-

EXPANSION PHASE: ENERGY DISSIPATION

netograms, and give a measure of the intensity of the

ionospheric currents in the auroral zone. The AL in-

dex is often used as a measure of substorm intensity.

Various studies have shown that the AL index is cor-

related with the interplanetary magnetic eld with

TAILWARD MOVING PLASMOID

t at ab out 20-min delay from the IMF an enhancemen

RECOVERY PHASE

southward turning and a larger intensi cation after

ab out 1 hour of southward IMF or enhanced energy

input Bargatze et al. 1985. The start of the sub-

storm growth phase at 0644 UT and the substorm wn with dotted lines in the onset at 0731 UT are sho

gure. The global auroral imagesdata not shown re-

veal the large-scale nature of the substorm pro cesses:

within only a few tens of minutes, a large p ortion of

Figure2. Schematic of the substorm sequence. The g-

the auroral ionosphere { and a large p ortion of the

ure shows the growth phase-associated thin current sheet

inner and mid-magnetotail { were in uenced by the

formation, plasmoid ejection, and recovery of the initial

energy dissipation pro cesses.

state. The top panel indicates scale sizes of the system

distance to the Moon is about 60 R .

The magnetic eld at geostationary orbit showed typ-

ical growth phase variations, decrease of the north-

Geomagnetic storms are large disturbances in the ward comp onent shortly after the IMF southward

near-Earth envoronment caused by coherent solar turning. Figure 4 shows the H comp onent of the

wind and interplanetary eld structures that origi- magnetic eld, which p oints northward along the

nate from solar disturbances such as coronal mass dip ole axis. GOES 9 at ab out 2300 MLTwas initially

ejections Gonzalez et al. 1994. Storms are asso ci- outside the disturbance region, and hence observed a

ated with ma jor disturbances in the in the geomag- further decrease of H from 0735 to 0740 UT, after 3

Dec 10, 1996 15 10 5 0 Dec 10, 1996 -5 100 -10 Bz [nT] WIND 80 0644 0731 0800

6 7 8 9 10 60 10 8 0644 0731 Epsilon [1011 W] 40 6 (Time shifted 7 min) 20 4 H [nT] GOES-8 and GOES-9

2 105 4 4 15 10 3 Integrated epsilon [10 J] 103 2 (Time shifted 7 min) 102 1 WIND 101 e- flux [1/cm2 sr s keV] (50-315 keV) 1994-084

0 -200 AL [nT] 105 -400 104 -600 3 -800 10 CANOPUS 2 10 4 1 6 10 3 Integrated -AL [10 nT s] p+ flux [1/cm2 sr s keV] (50--670 keV) 1994-084

2 6 7 8 9 10 UT 1 CANOPUS

6 7 8 9 10

Figure 3. Substorm driver and its e ects in the iono- Figure 4. Inner magnetosphere e ects of substorm ac-

sphere. Interplanetary magnetic eld B from WIND tivity. Magnetic eld H component from GOES-9 and

MFI instrument. The parameter and dt computed GOES-8 spacecraft. Energetic electron energy range 50{

using the solar wind velocity and magnetic eld from 315 keV and proton energy range 50{670 keV mea-

WIND. AL index from CANOPUS and its time integral surements from s/c 1994-084. GOES data courtesy of H.

Singer; LANL data courtesy of G. Reeves. ALdt. From Pulkkinen et al. 1998b.

3. GEOMAGNETIC STORMS: May 15, 1997

which the eld gradually b ecame more dip olar; a -

An earthward directed, large coronal mass ejec-

nal, stronger dip olarization o ccurred after 0810 UT.

tion CME was observed on May 12, 1997 by the

GOES 8 in the p ostmidnight sector was to the east of

SOHO Extreme ultraviolet Imaging Telescop e EIT

the initial activity region and showed only one grad-

at ab out 0450 UT data not shown, see Thompson

ual dip olarization when the currentwedge expanded

et al. 1998. The CME created a disturbance More-

over the satellite longitude from 0842 UT onward. At

ton wave, which in the 195 A di erence images was

1500 MLT at geostationary orbit, the Los Alamos Na-

seen to propagate over the entire visible solar disk.

tional Lab oratory energetic particle instrument de-

A few hours later, the CME was observed by the

tected enhancements in b oth proton 50{670 keV

Large Angle Sp ectrometric Coronagraph LASCO

and electron 50{315 keV uxes. These enhance-

as a halo event; a bright expanding ring centered

ments app eared later than the actual substorm activ-

around the o cculting disk Brueckner et al. 1998.

ity, as the satellite was far from the substorm onset

region.

The coronal mass ejection expanded from the Sun

and was later observed by the WIND spacecraft up-

stream of the Earth as a magnetic cloud Baker et al.

1998b. Figure 5 shows the WIND measurements of

The signatures intro duced here, enhanced energy in-

the total magnetic eld, the B comp onent of the

put caused by southward IMF; enhancement of auro-

eld, solar wind velo city and density, and the pa-

ral luminosity and electro jets; stretching of the mag-

rameter and its time integral. Note how the strongly

netic eld followed by a dip olarization; and enhance-

southward IMF causes an almost order of magni-

ment of energetic tens to hundreds of keV electrons

tude larger p eak energy input value during the

and protons near geostationary orbit are all typical

cloud interval. Furthermore, the integrated energy

signatures of the substorm evolution, and their tem-

input dt during the cloud was a factor of ve

p oral sequence is well-established. The asp ect that

larger than that asso ciated with the large substorm

makes substorms dicult in terms of forecasting is

describ ed in the previous section.

that the relationship b etween the energy input, stor-

age, and dissipation is highly nonlinear and hence

very sensitive to the initial conditions see e.g. Baker As the cloud reached the Earth, the magnetosphere

et al. 1996a. Several attempts to predict e.g. the resp onded with an increase of the level of activityas

AL index from upstream data have b een made b oth recorded by several parameters. The auroral zone

using nonlinear prediction lters and neural networks magnetograms CU/CL indices denoting the lo cal

Gleisner & Lundstedt 1997, Klimas et al. 1998a, AU/AL indices from the Canadian lo cal time sector

but the prediction accuracy b oth in terms of intensity showed strong auroral activity reaching 1500 nT, and

and timing still require re nements in the mo dels. the geostationary orbit data show large disturbance

crease in the relativistic electron uxes was asso ci-

ated with the strong substorm activity and the p ersis-

twave activity present in the inner magnetosphere May 1997: WIND ten

er et al. 1998b. The electron uxes increased by 25 Bak

20 B [nT]

wo orders of magnitude, and p ersisted at the 15 almost t

Magnetic ated level for several days. 10 elev 5 cloud interval 0 20 10 0 -10 -20 B [nT] May 14-17, 1997 -30 Z 700 200 600 V [km/s] 150 500 100 Appearance of 400 50 B [nT]relativistic e- GOES-8 300 0 TOT 200 150 50 -3 40 n [cm ] 100 30 50 20 0 B [nT] GOES-8 10 -50 X 0 1000.0 100.0 e- flux (>2 MeV) [cm-2 s-1 sr-1] GOES-8 11 100.0 10.0 epsilon [10 W] 10.0 1.0 1.0 0.1 0.1 106 20 -2 -1 -1 Integrated epsilon [1016 J] 5 e- flux [cm s sr ] LANL 15 10 4 10 10 3 5 10 102 0 50 May 14 May 15 May 16 May 17 0 Days of 1997 -50 -100 -150 Dst [nT] -200 500 CU 0 CL -500 -1000

Canopus

e5. Magnetic cloud in the interplanetary medium.

Figur -1500

WIND measurements of the total magnetic eld, B , so-

Z May 14 May 15 May 16 May 17 R

Days of 1997

lar wind velocity, density, and dt computed from

the solar wind and IMF parameters. From Baker et al.

1998b.

Figure 6. E ects of the magnetic cloud in the magne-

levels with several substorm injection events. After

tosphere and ionosphere. Total magnetic eld and B

the cloud arrival, during the storm main phase, the

as measured by GOES-8, relativistic electron ux from

electron uxes were extremely variable, re ecting the

GOES-8, energetic electron ux from Los Alamos instru-

strong disturbance level of the entire inner magneto-

ments, and ground-based Dst and CU/CL indices. From

sphere Figure 6.

Baker et al. 1998b.

The Dst-index reached a minimum of 169 nT

Baker et al. 1998b. The Dst index is a measure

Measurements from SAMPEX on a p olar orbit at 800

of the intensity of the ring current carried by ener-

km altitude provide another way to examine the rel-

getic ions in the inner magnetosphere L 4 7.

ativistic electron resp onse to the storm activity. Fig-

During the storm main phase the ring currentinten-

ure 7 shows an L vs time plot of 2{6 MeV electron

si es continuously and hence Dst decreases; the slow

uxes throughout the year of 1997. Several inten-

recovery that takes from one to several days is asso-

si cations of the electron uxes are clearly visible,

ciated with the life time of the ring current ions. The

mayb e the most notable b eing the one at the b egin-

Dst index is commonly used as a measure of the ge-

ning of the year related to the geomagnetic storm on

omagnetic storm intensity. Both neural network and

Jan 10, 1997 Baker et al. 1998a, Reeves et al. 1998.

nonlinear prediction lter studies relating solar wind

The intensi cation on May 15, 1997 is distinguished

parameters and Dst have shown that using L1 mea-

by the large intensity enhancementaswell as its low

surements and previous values, Dst can b e predicted

L-shell; ux enhancements extended down to the slot

one hour ahead with quite go o d level of accuracy Wu

region at L<3between the inner and outer radiation

& Lundstedt 1997, Klimas et al. 1998b.

b elts.

The magnetic eld at geostationary orbit showed

strong uctuations during the entire p erio d of south-

4. DISCUSSION

ward IMF. Furthermore, detailed analysis of the au-

roral zone ground magnetograms reveals that there

was strong ULF wave activity in the frequency range

4.1. Observational Needs

of 2{20 mHz data not shown; see Baker et al.

1998b. Near the p eak of the storm main phase,

a rapid increase in the relativistic electron uxes was The key element in all space weather prediction and

observed at the GOES-8 spacecraft. The rapid in- forecasting activities is continuous monitoring of the

needs to consider which solar observations are most

critical for the forecasting purp oses and what kind of E orts should al- 8 Flux monitors are required in the future.

10000.0

ready now b e fo cussed to problems of how to predict

activity from much less detailed data, and what the

um requirements for solar observations would 1000.0 minim

6 be.

5 100.0

Lshell 10.0

4.2. Metho ds for Predicting Substorms and Storms 3

1.0

2 May 15, 1997

4.2.1. Solar Observations SAMPEX/ELO E = 2-6 MeV 1 0.1 0 100 200 300

Time [days of 1997]

The b est indicators that a coronal mass ejection is

on its waytoward the Earth are a brightening in the

EIT images on the solar disk followed by a halo event

in the coronagraph LASCO observations. How-

ever, Brueckner et al. 1998 studied the correlation

Figure7. E ects of the magnetic cloud: L-sorted electron

of CMEs and geomagnetic activity, and concluded

uxes measured at low altitudes by SAMPEX/ELO for

that also other kinds of CMEs than the halo events

2{6 MeV. The horizontal axis shows the day of year, the

needed to b e included to explain all storms that o c-

vertical axis shows the magnetic L-shel l, and the electron

curred during the p erio d of investigation. They also

ux is grey-scale codedaccording to the scale on the right.

concluded that usually it to ok ab out 80 hours for the

Data courtesy of B. Klecker.

CME to reach the Earth, with a few exceptions with

longer delays. Such expansion sp eeds are consistent

with a ux rop e mo del of CME expansion Chen et al.

Sun and the upstream solar wind. Presently NASA

1997, Chen 1996. These mo dels provide a go o d basis

op erates two scienti c spacecraft WIND and ACE,

for mo deling the CME propagation from the Sun to

which provide almost continuous coverage of solar

the Earth in order to predict storm o ccurrence based

wind parameters and the interplanetary magnetic

on solar observations.

eld at or close to the L1-p oint. Measurements made

at L1 give ab out 1 hour warning time under typical

solar wind sp eeds, and provide a reasonably accu-

rate estimate of the plasma and eld prop erties in

4.2.2. Physical Mo dels

the vicinity of the Earth. However, di erences in the

solar wind and IMF front orientations can pro duce

substantial di erences in the travel times from L1 to

The global MHD simulations provide at present the

Earth, and if the spacecraft is away from the Sun-

only means to physically mo del the e ects of solar

Earth line, smaller disturbances may not reach the

wind and IMF variations on the large-scale magneto-

magnetosphere Ridley et al. 1998. Hence, two or

sphere { ionosphere system. These mo dels accept the

more spacecraft in the upstream solar wind at dif-

solar wind and IMF parameters as input, and then

ferent distances would b e very advantageous for the

compute the time evolution of the coupled solar wind

forecasting.

{ magnetosphere { ionosphere system using MHD-

equations Janhunen et al. 1996. At present, there

It is also imp ortant to notice that b oth ACE and

are ab out ten such simulation co des in the world, one

WIND are scienti c satellites, which do not have im-

of which in Europ e Bourdarie 1998. The mo del de-

mediate backups if failures would o ccur. Therefore, a

velopment and computer resources are at the stage

key issue for space weather forecasting in the future

where the results are accurate enough to b e compared

is that one or more of the large space agencies would

with actual measurements Pulkkinen et al. 1998b

take resp onsibility for providing continuous series of

and the computations are fast enough to be run in

simple solar wind monitors and their uninterrupted

real time. It is to b e exp ected that these mo dels will

tracking.

advance substantially toward op erational use during

the next few years. Such mo dels can b e used equally

well for the large disturbances during geomagnetic

At present, SOHO provides an excellent means to

storms as well as the smaller substorm events.

monitor the solar activity in great detail. The aim

of the ISTP program is to gain b etter understanding

of the Sun-Earth connection, of the chain of events

However, a key factor for space weather is the en-

from the surface of the Sun through the magneto-

ergetic particle environment in the inner magneto-

sphere, ionosphere, and upp er atmosphere. The next

sphere. For MHD simulations, this p oses two imp or-

few years during the solar maximum will most prob-

tant problems: First, the mo dels are valid only to 3.5

ably bring ma jor steps forward in our capability to

R from the Earth, and hence some b oundary e ects

relate the Solar and magnetospheric pro cesses with

can o ccur in the inner magnetosphere. Furthermore,

each other. While unambiguous asso ciation of solar

the MHD theory describ es the entire plasma with

and magnetospheric disturbances is more dicult un-

a single temp erature, and hence do es not provide a

der active conditions, suchwork is of key imp ortance

means to examine the high-energy particle dynam-

for space weather predictions Brueckner et al. 1998,

ics. This environment has to b e examined using sp e-

Bothmer & Schwenn 1995.

ci c mo dels for the radiation b elts, which can use

the MHD magnetic and electric elds as mo del input

Bourdarie 1997.

However, after SOHO the space weather community

4.2.3. Data-derived Mo dels and from the enhanced understanding of the under-

lying physical pro cesses.

The nonlinear nature of the solar wind { magne-

tosphere coupling makes predicting substorm activ-

ACKNOWLEDGMENTS

ity dicult. Several authors have studied this cou-

pling using prediction lters and neural network tech-

niques to deduce the geomagnetic indices based only

The author would like to thank K. Ogilvie and R.

on upstream solar wind measurements and previous

Lepping for the use of WIND data in this pap er, the

time history of the index. These metho ds have given

Canadian Space Agency and the CANOPUS team for

promising results b oth on AL index substorm timing

the use of CANOPUS magnetometer data, G. Reeves

and intensity, Klimas et al. 1998a, Gleisner & Lund-

for the LANL energetic particle measurements, H.

stedt 1997 and Dst storm timing and intensity, Kli-

Singer for the GOES magnetic eld and electron

mas et al. 1998b,Wu & Lundstedt 1997 predictions.

data, and B. Klecker for the SAMPEX data.

Usually, the indices are calculated one hour in ad-

vance, and the predictions use the measured values

from the previous hour, the solar wind and IMF data

as input.

REFERENCES

Acuna, ~ M. H., K. W. Ogilvie, D. N. Baker, S. A. Cur-

4.3. Di erences Between Storms and Substorms

tis, D. H. Fair eld, and W. H. Mish, The global

geospace science program and its investigations, in

The Global Geospace Mission, edited by C. T. Rus-

Because there are signi cant di erences b etween the

sell, p. 5, Kluwer Acad., Dordrecht, the Nether-

storm and substorm dynamics, the prediction of

lands, 1995.

storms and substorms are quite di erent problems.

Baker, D. N., T. I. Pulkkinen, V. Angelop oulos, W.

Baumjohann, and R. L. McPherron, The neutral

Geomagnetic storms are driven by fast solar wind

line mo del of substorms: Past results and present

and/or strong and southward interplanetary mag-

view, J. Geophys. Res., 101, 12,975, 1996a.

netic eld. Such solar wind/IMF conditions are a

consequence of an activeevent on the Sun. The mag-

Baker, D. N., J. H. Allen, R. D. Belian, J. B.

netospheric disturbances are directly driven by the

Blake, S. G. Kanekal, B. Klecker, R. P. Lep-

solar wind conditions and last for 1{2 days. Storm-

ping, X. Li, R. A. Mewaldt, K. Ogilvie, T. On-

time disturbances accelerate electrons often to MeV

sager, G. D. Reeves, G. Rostoker, H. J. Singer,

energies.

H. E. Sp ence, and N. Turner, An assessment

of space environmental conditions durin the re-

Substorms are smaller disturbances that last 2{4

cent Anik E1 spacecraft op erational failure, ISTP

hours and are driven by southward p erio ds of the

Newsletter, 6, No 2, June, 1996b http://www-

interplanetary magnetic eld. Suchvariations are al-

istp.gsfc.nasa.gov/istp/newsletter.html.

ways present in the solar wind; hence substorms o c-

Baker, D. N., T. I. Pulkkinen, X. Li, S. G. Kanekal,

cur at a rate of several p er day regardless of the so-

J. B. Blake, R. S. Selesnick, M. G. Henderson, G.

lar activity although substorms are more numerous

D. Reeves, and H. E. Sp ence, Coronal mass ejec-

and stronger during solar maximum than during so-

tions, magnetic clouds, and relativistic magneto-

lar minimum Nevanlinna & Pulkkinen 1998. While

spheric electron events: ISTP, J. Geophys. Res.,

storms are directly driven by the solar wind and IMF,

103, 17279-17292, 1998a.

substorms include also an energy unloading pro cess

Baker, D. N., T. I. Pulkkinen, X. Li, S. G. Kanekal,

governed byinternal magnetospheric pro cesses. Dur-

K. W. Ogilvie, R. P. Lepping, J. B. Blake, L.

ing substorms, particles b oth electrons and protons

B. Callis, G. Rostoker, H. J. Singer, and G. D.

are typically accelerated to few hundred keV ener-

Reeves, A strong CME-related magnetic cloud in-

gies. During storms, the particles accelerated by sub-

teraction with the Earth's magnetosphere: ISTP

storms form the seed p opulation that is then further

observations of rapid relativistic electron accelera-

accelerated to the MeV energies Baker et al. 1998a.

tion on May 15, 1997, Geophys. Res. Lett. 25, 2975,

1998b.

Bargatze, L. F., D. N. Baker, R. L. McPherron,

4.4. Solar Minimum and Solar Maximum

and E. W. Hones, Jr., Magnetospheric impulse re-

sp onse for many levels of geomagnetic activity, J.

The Sun is vastly more active during solar max-

Geophys. Res., 90, 6387, 1985.

imum than it is during solar minimum, but geo-

Brueckner, G. E., J.-P. Delab oudiniere, R. A.

magnetic storms and esp ecially substorms o ccur also

Howard, S. E. Paswaters, O. C. St.Cyr, R.

during minimum p erio ds. Notice that b oth of the

Schwenn, P. Lamy, G. M. Simnett, B. Thomp-

events presented in this pap er are from the mini-

son, and D. Wang, Geomagnetic storms caused

mum p erio d. Therefore, as technology is and con-

by coronal mass ejections CMEs: March 1996

tinues to be susceptible to problems caused by the

through June 1997, Geophys. Res. Lett., 25, 3019,

plasma environment, space weather as an issue will

1998.

not disapp ear with the next solar maximum. The

Bothmer, V., and R. Schwenn, The interplanetary

space physics community is fo cusing great e orts to

and solar causes of ma jor geomagnetic storms, J.

monitor, mo del, and develop theories for the active

Geomag. Geoelectr., 47, 1127, 1995.

events that o ccur during the up coming solar maxi-

mum. The space weather applications will certainly Bourdarie, S., D. Boscher, T. Beutier, J.-A. Sauvaud,

b ene t greatly b oth from the monitoring capabilities and M. Blanc, Electron and proton radiation b elt

dynamic simulations during storm p erio ds: a new Ridley, A. J., G. Lu, C. R. Clauer, and V. O. Papi-

asymmetric convection{di usion mo del, J. Geo- tashvili, A statistical study of the ionospheric con-

phys. Res., 102, 17541, 1997. vection resp onse to changing interplanetary mag-

netic eld conditions using the assimilative map-

Bourdarie, S., Physical magnetospheric and radiation

ping of ionospheric electro dynamics technique, J.

b elt mo dels MHD and di usion co des, this issue.

Geophys. Res., 103, 4023, 1998.

Chen, J., Theory of prominence eruption and prop-

Rostoker, G., S.-I. Akasofu, J. Foster, R. A. Green-

agation: interplanetary consequences, J. Geophys.

wald, Y. Kamide, K. Kawasaki, A. T. Y. Lui, R.

Res., 101, 27499, 1996.

L. McPherron, and C. T. Russell, Magnetospheric

Chen, J., R. A. Howard, G. E. Brueckner, R. San-

substorms - de nitions and signatures, J. Geophys.

toro, J. Krall, S. E. Paswaters, O. C. St.Cyr, R.

Res., 85, 1663, 1980.

Schwenn, P. Lamy, and G. M. Simnett, Evidence

Thompson, B. J., S. P. Plunkett, J. B. Gurman, J.

of an erupting magnetic ux rop e: LASCO coro-

S. Newmark, O. C. St.Cyr, and D. J. Michels,

nal mass ejection of 1997 April 13, Ap. J., 490,

SOHO/EIT observations of an Earth-directed

L191, 1997.

coronal mass ejection on May 12, 1997, Geophys.

Gleisner, H., and H. Lundstedt, Resp onse of the au-

Res. Lett., 25, 2465, 1998.

roral electro jets to the solar wind mo deled with

Wu, J.-G. and H. Lundstedt, Geomagnetic storm

neural networks, J. Geophys. Res., 102, 14269,

predictions from solar wind data with the use of

1997.

dynamic neural networks, J. Geophys. Res., 102,

Gonzalez, W. D., J. A. Joselyn, Y. Kamide, H. W.

14255, 1997.

Kro ehl, G. Rostoker, B. T. Tsurutani, and V. M.

Vasyliunas, What is a geomagnetic storm?, J. Geo-

phys. Res., 99, 5771, 1994.

Janhunen, P., H. E. J. Koskinen, and T. I. Pulkkinen,

A new global ionosphere-magnetosphere coupling

simulation utilizing lo cally varying time step, in:

Third International Conference on Substorms ICS-

3, ESA-SP 389, p. 205, 1996.

Klimas, A. J., J. A. Valdavia, D. Vassiliadis, and D.

N. Baker, AL index prediction using data-derived

nonlinear prediction lters, Phys. Space Plasmas 15,

1998a.

Klimas, A. J., D. Vassiliadis, and D. N. Baker, Dst in-

dex prediction using data-derived analogues of the

magnetospheric dynamics, J. Geophys. Res., 103,

20435, 1998b.

McPherron, R. L., Physical pro cesses pro ducing mag-

netospheric substorms and magnetic storms, in

Geomagnetism,vol. 4, p. 593, Academic Press, San

Diego, CA, 1991.

Nevanlinna,H., and T. I. Pulkkinen, Solar cycle cor-

relations of substorm and auroral o ccurrence fre-

quencies, Geophys. Res. Lett., 25, 3087, 1998.

Perreault, P., and S.-I. Akasofu, A study of geomag-

netic storms, Geophys. J. R. Astr. Soc., 54, 547,

1978.

Pulkkinen, T. I., D. N. Baker, L. L. Cogger, T.

Mukai, H. J. Singer, Coupling of inner tail and

midtail pro cesses, in: Substorms-4, edited by S.

Kokubun and Y. Kamide, Terra Scienti c Pub-

lishing Company,Tokyo, p. 749, 1998a.

Pulkkinen, T. I., D. N. Baker, M. Wiltb erger, C.

Go o drich, J. G. Lyon, and R. E. Lop ez, Pseu-

dobreakup and substorm onset: Observations and

MHD simulations compared, J. Geophys. Res., 103,

14847, 1998b.

Pulkkinen,T. I., D. N. Baker, L. L. Cogger, L. A.

Frank, J. B. Sigwarth, S. Kokubun, T. Mukai,

H. J. Singer, K. Ogilvie, J. Slavin, and L. Ze-

lenyi, Spatial extent and dynamics of a thin cur-

rent sheet during the substorm growth phase on

Dec 10, 1996, J. Geophys. Res., submitted, 1998c.

Reeves, G. D., R. H. W. Friedel, M. G. Henderson, R.

D. Belian, M. M. Meier, D. N. Baker, T. Onsager,

and H. J. Singer, The relativistic electron resp onse

at geosynchronous orbit during the January 1997

magnetic storm, J. Geophys. Res., 103, 17559, 1998.