The auroral oval polar boundary position


of Atmosphcrrc

and Solar-Trrratrrol

Pergamon PII: SOO21-9169(96)00009-8

Physics, Vol CopyrIght

59, No. 2, pp. 23 l-235. 0 1996 Elsevier Scmce

1997 Ltd

Printed in Great Britain All rights reserved 136&6826/97$17.00+0.00

Rapid Communication The aurora1 oval polar boundary position E. V. Voronov’


S. V. Fridman’*

of Solar-Terrestrial Physics, 664033 , Irkutsk, P.O. Box 4026, Russia; *University Urbana-Champaign, 1308 West Main Street, Urbana, IL 61801-2307, U.S.A.


(Received injinalform

3 October

of Illinois at

1994; accepted 10 October 1994)

is considered that a part of the polar cap, adjacent to the aurora1 oval, lies on closed magnetic field lines. This has found a reasonable explanation in the framework of a model of the plasma sheet as a collisionless shock wave (Fridman and Voronov, 1993, J. geophys. Res. A98, 143-151). According to the model, the polar cap can exist even if all of its magnetic field lines intersect the plasma sheet. The global magnetospheric electric field controls the dynamics of the aurora1 oval polar boundary. An abrupt enhancement of the global magnetospheric convection electric field is followed immediately by a thinning of the plasma sheet and by an equatorward displacement of the boundary. With a decrease of the magnetospheric electric field, the boundary moves toward the cusp. Whether the electric field increases or decreases, the local aurora1 formations inside the aurora1 oval move equatorward with respect to the Earth. Copyright c 1996 Elsevier Science Ltd


is difficult to explain the existence of closed magnetic lines in the polar cap. On the other hand, this fact looks quite natural in the framework of a concept of the magnetospheric plasma sheet as a collisionless shock wave as suggested recently by Fridman and Voronov (1993). In that paper we hypothesized the plasma sheet to be a hot plasma region between two collisionless compression shock waves that have moved away from the neutral plane. The shock waves are formed as a result of the convection of closed magnetic tubes in the dawn-dusk electric field that reduce their length. It was not a self-consistent solution of the problem: we only evaluated the geophysical consequences of the proposed hypothesis. According to this concept, the plasma sheet boundary is formed as a result of the competition between three motions: (1) the motion of the shock front with respect to the cold background plasma of the tail lobes along a magnetic field line in the direction from the neutral plane; (2) the opposing motion of the background plasma; and (3) the E x Bdrift that usually is directed towards the neutral sheet. This results in the formation of the plasma sheet, whose meridional cross-section is shown in Fig. la. Such a model explains quite well a large number of known features of the real plasma sheet and, amongst other things, the existence of its clear-cut boundary with a small thickness and a large abrupt change in plasma density. In terms of this model we interpret




and polar

of many


problems physics


of the polar

field topology



as a region









on the correct

cap size and

the magnetic

it. The polar that

in magnetospheric


cap is defined

lies poleward it was

the region polar





is determined cap


of the

by the mag-




are open, but aurora1 oval field lines are closed and projected onto the plasma sheet and its boundary layer (Lyons and Williams, 1984). Recent experimental evidence suggests that the portion of the polar cap just inside the aurora1 oval, lies on closed magnetic field lines. For example, consider the observation of the socalled ‘double’ aurora1 oval (Elphinstone and Hearn, 1992; Murphree and Cogger, 1992). This distribution consists of the main aurora1 oval and another parallel aurora1 band 5-10” poleward of it. Airglow is caused by particles precipitating from the plasma sheet. If it is assumed that the particle distributions in the plasma sheet are near-isotropic, the plasma completely fills the magnetic flux tubes, and the precipitation comes from the tubes’ feet resting on the atmosphere, then it

*On leave from Irkutsk, Russia.


of Solar-Terrestrial

Physics, 231


E. V. Voronov

and S. V. Fridman

6 LT

h) -~

0 LT

Fig. 1. (a) Plasma sheet cross-section in the meridional plane; (b) the intersection of the shock surface with the ionospheric level (the polar boundary of aurora1 oval). The magnetic field model (1) has been used. We have set the electric field value E, = 0.15 mV/m. Magnetic field line bending on the plasma sheet boundary (i.e. on the shock surface) is not shown.

the line of ionospheric oval (Fig. etrate into sions only


of the shock




level as the polar boundary of the aurora1 1b) because plasma sheet particles can penthe atmosphere and cause aurora1 emisequatorward of this boundary.




An important feature of the model suggested here is that the plasma sheet boundary does not coincide with any of the magnetic surfaces, but is intersected by magnetic field lines. In each of the tail lobes there exists a region occupied by closed field lines but not containing hot plasma. This means that the spatial location of the polar cap boundary seems to be associ-

field topology to a lesser extent than is believed usually. Quite a realistic polar cap will form even if the magnetic field does not include open field lines at all. To illustrate this, when constructing Fig. 1, we have used a very simple magnetic field model, given in the solar-magnetospheric coordinate system by ated with the magnetic

B = & + (& tanh (z/d), 0, &),


where B, is the dipole field, B,, and B:, are the Xand z-components of the geomagnetic field in the tail lobes, respectively, tanh is the hyperbolic tangent, and d is the parameter that determines the current sheet thickness. In spite of its evident weaknesses, model (1) has an important feature for the illustration of features

The aurora1 oval polar boundary position of our model, namely all of its magnetic field lines are closed. We consider the electric field in the neutral plane of the tail to be uniform with strength E,,, and directed dawn-dusk. We also consider the magnetic field lines to be equipotentials, and so we map the electric field from the tail along them. Specifically, we have adopted B,, = 8 nT, Bra = 2 nT, d = 2 R,, and Em = 0.15 mV/m z 1 kV/RE (RE being the Earth’s radius). According to this hypothesis, the E x B-drift is one of the decisive factors responsible for the plasma sheet configuration; hence, any variation in the global convention electric field must have a direct effect on this configuration and on the location of the polar boundary of the aurora1 oval. To put it another way, this boundary motion is not unambiguously determined by reconfiguration of the magnetic field. Moreover, it can occur even though the magnetic field does not change at all. Of course, abrupt changes in the magnetospheric electric field pattern seem to be part of the processes causing substantial rearrangement of the magnetic field. Nevertheless, we deliberately neglect this rearrangement with the exception of the magnetic field B bending at the shock front in order to derive the effect of electric field variations in a ‘pure form’. Let us consider the dynamics of the plasma sheet boundary in the tail when the magnetospheric electric field changes. It is easy to understand that, since the plasma sheet boundary position is determined by the competition of field-aligned shock front motion and E x B-drift, the electric field enhancement leads to plasma sheet thinning and an equatorward displacement of the aurora1 oval polar boundary. Our model predictions on this point are the same as the results of works in which a plasma sheet boundary is considered to be formed by plasma streams originating from the region in the vicinity of the distant tail neutral line (Hill, 1975; Cowley, 1980; Owen and Cowley, 1987 etc.). But there is a sufficient distinction between such models and ours: in our model we do not invoke the hypothesis about global magnetic reconnection at all. Magnetic reconnection is not important in our model; what matters here is the presence of tail-to-Earth plasma convection only. Let us use the most simple kind of magnetospheric electric field variation, namely an abrupt change of its value from E,,,, to Em2 at time t = 0. By abrupt, we of course mean on a time scale longer than the typical time of processes in the shock wave (7). This change can be estimated roughly as the ratio of the shock front thickness AZ (i.e. the plasma sheet boundary layer thickness) to the thermal velocity of the ions: 7 = Az/(2 w/m)I!‘. In the tail where the typical energy





20 40 60 80 100 Time, minutes Fig. 2. The displacement of the polar boundary of nightside aurora1 oval after an abrupt change of the electric field value from E,,, to E,, at time t = 0. Lines l-3 correspond to the following cases- (1) the enhancement of electric field from to Em7 E_, ,I . = 0.45 mV/m, s3 ,,..=0.15 mV/m, x 1 kV/R,, kV/R,; (2) the electric field decreases from E,,,, = 0.15 mV/m to &I, = 0; (3) convection reversal: E,, = 0.15 mV/m, Em2= -0.15 mV/m.

of the protons is W z 1 keV and AZ = 0.1-0.2 RE, this time is 7 = 1.5-3 s. Figure 2 presents the numerically calculated location of the nightside aurora1 oval polar boundary. Line 1 seems to represent the most simple case: the enhancement of electric field from the value EmI = 0.15 mV/m, =: 1 kV/Rs, typical of quiet conditions, to E,, = 0.45 mV/m, z 3 kV/RE. The nightside polar cap boundary moves equatorward by about 3” in a time 7 , z 40 min, which is the time of convection between the stationary L-shells occupied by the ‘old’ and ‘new’ polar boundaries of the region of precipitation from the plasma sheet. In the case of a northward IMF, a long-duration significant decrease of the magnetospheric electric field is possible; let it be specified as Em, = 0.15 mV/m, Em2 = 0. In this case the E x B-drift of plasma ceases, but the shock wave continues to move along a magnetic field line, causing the plasma sheet to expand and the aurora1 oval to contract toward the cusp (line 2). In principle, all closed field lines of the polar cap can be filled with plasma sheet particles, the characteristic time of this expansion being several hours. A short-time convection reversal becomes possible with a northward IMF: E,,, = 0.15 mV/m, E,, = -0.15 mV/m. In this case (line 3) the electric drift favours, rather than hinders, the expansion of the plasma sheet; if the reversed convection exists during


E. V. Voronov

and S. V. Fridman

a time interval longer than TV= R,B,,/cE,,,,, where R, z 20 RE is the tail radius, then the plasma sheet can expand to reach the size of the entire tail and weak aurora1 can cover the former polar cap fully. For typical values of the magnetospheric parameters (Em zz 0.15 mV/m, B, E 10 nT) we have 72 z 8 * lo3 s z 2 h, but thickening of the plasma sheet can become significant during a substantially shorter time interval. As can be seen, the processes in the magnetotail and in the ionosphere look like the processes of the previous case, but are significantly faster. An expansion of the plasma sheet is accompanied by a significant decrease in density of the contained plasma, and so both the plasma sheet and aurora1 oval may be thought of as disappearing. It should be noted that such a great thickening of the plasma sheet and auroral oval polar boundary displacement arise only in the extremely simplified magnetic field model (1) used here. The real magnetotail can contain both closed and open magnetic lines and, moreover, would be subject to magnetic field reconfiguration. So the plasma sheet thickening appears to capture only the region of closed magnetic lines. In general, the motion of the polar boundary of the aurora1 oval does not coincide with the motion of aurora1 formations inside the oval. It has been shown that the model polar boundary moves poleward or equatorward when the cross-tail electric field decreases or increases, respectively. In contrast to this, if the convection retains its former direction, local aurora1 formations inside the aurora1 oval will always move equatorward with respect to the Earth as they are caused by plasma on contracting field lines. It is important to note that the polar boundary of the oval starts to move immediately after an enhancement of the electric field. This does not contradict the fact that inside the aurora1 oval brightenings can manifest themselves only in a ‘transport time’ of about l-l.5 h,

which is associated with the arrival of part of the dense hot plasma from the distant tail. For example, the substorm Sauvaud et al. (1987) by investigating of 4 March 1979 on the basis of simultaneous observations in the solar wind, at geosynchronous orbit and on the ground, found that the equatorward displacement of the auroras started with an insignificant delay (5-10 min) following the southward turning of the IMF vector, whereas the main activation of the auroras did not set in until 80 min after this turning. A statistical analysis made by Weimer et al. (1992) showed that the electric potential difference across the polar cap begins to grow 1.5 h before the beginning of the substorm expansion phase. The polar cap also begins to grow in size (from 24” to 31”) about 1 h before the expansion phase.


There are closed magnetic lines of the polar cap which are adjacent to the aurora1 oval. Moreover, the polar cap can exist, even though all of its magnetic lines intersect the plasma sheet. An enhancement of the magnetospheric convection electric field causes a displacement of the aurora1 oval polar boundary equatorwards, and with a decrease of it the boundary is pulled closer to the cusp. Whether the dawn to dusk directed electric field increases or decreases, the local aurora1 formations inside the aurora1 oval move equatorward. The above motions set in immediately following a change in the global electric field. Acknowledgements-This research was supported by the Russian Foundation for Fundamental Researches under Grant 94-05-16167-a. We are indebted to V. G. Mikhalkovsky for his assistance in preparing the English version of the manuscript.



Cowley S. W. H. Elphinstone


R. D. and Hearn

S. V. and Voronov

D. J.

E. V.



Hill T. W. Lyons L. R. and Williams Murphree


J. S. and Cogger

D. J. L. L.

1984 1992

Plasma populations in a simple open model magnetosphere. Space Sci. Rev. 26.217-275. Mapping of the-aurora1 distribution during quiet times and substonn recovery. Proc. International Conference on Substorms (ICS-l), Kiruna, Sweden, 2327 March 1992, ESA SP-335, pp 13-18. The plasma sheet as a shock wave: geophysical consequences of the hypothesis. J. geophys. Res. A98,143151. Magnetic merging in a collisionless plasma. J. geophys. Res. 80,468994699 Quantitative Aspects of Magnetospheric Physics. D. Reidel, Dordrecht, 1984. 231 pp. Observations of substorm onset. Proc. International Conference on Substorms (ICS-l), Kiruna, Sweden, 23-27 March 1992, ESA SP-335, pp. 2077211.

The amoral

oval polar boundary

Owen C. J. and Cowley S. W. H.


Sauvaud J. A., Treilhou J. P., Saint-Marc A., Dandouras J., Reme H., Korth A., Kremser G., Parks G. K., Zaitzev A. N., Petrov V., Lazutine and Pellinen R. Weimer R. D.. Kan J. R. and Akasofu S.-I.

1987 L 1992



Simple models of time-dependent reconnection in a collision-free plasma with an application to substorms in the geomagnetic tail. Planet. space Sci. 35, 451l 466. Large scale response of the magnetosphere to a southward turning of the interplanetary magnetic field. J. geophys. Res. A92,2365-2376. Variations of the polar cap potential measured during magnetospheric substorms. J. geophys. Res. A97, 3945-3951.