Pkmet. Space Sci. 1970, Vol. 18, PP. 501 to 508. Perttamon Press. printed In Northern &dand
THE AURORAL OVAL AND THE BOUNDARY OF CLOSED FIELD LINES OF GEO~AG~TrC FIELD Y. I. F’ELDSTEIN Institut Terr. Magnetism, Ionosphere, Radio Wave Propagation, Academic Sciences U.S.S.R., Moscow, U.S.S.R.
G. V. STARKOV Polar Geophysical Institut, Academic Sciences U.S.S.R., Murmansk, U.S.S.R.
Abstract-A comparison of the position of the auroral oval with the boundary of the stable trapping region C&and the limit of closed geomagnetic field lines (4, has been carried out; Alouette-2 data are used to obtain the trapping boundary. In the midnight hours (5, coincides with equatorward boundary of the aurora1 ovaf, and in the midday hours & is situated within the oval. The equato~~d boundary of the aurora1 oval is closely connected with the position of the region, in which the geomagnetic field lines are closed, regardless of the degree of magnetic activity. The values of #, on the day of the Earth changes with universal time. It is suggested that the change is caused by the variation of the orientation of geomagnetic axis with respect to the streaming solar wind around the magnetosphere.
Piddington (1965), Akasofu (1966) and Fefdstein (1967) showed that the average position of the aurora1 oval coincided with the poleward border of the region where trapped electrons with E > 40 keV were observed. A comparison of these two phenomena was carried out for the periods with different levels of magnetic activity. However, it is known that the auroral oval (Feldstein et al., 1967) has considerable spatial variations with changes of magnetic disturbances and therefore the comparison of average values may be regarded as very approximate. In this paper an attempt is made to carry out a more detailed investigation of mutual position of borders of closed field lines and the auroral oval. McDiarmid et al. (1968a) determined the position of the border of closed field lines of the geomagnetic field for the period from December 1965 to June 1966. It was supposed that the border coincided with the region where the counting rate of the electrons with B > 35 keV fell to the cosmic-ray background level. In this observational period the ma~etic field was p~do~nantly quiet or only weakly distend. In more than 50 per cent of the cases the magnetic disturbance was characterized by the planetary K index K, = O-l; moreover the periods during geomagnetic storms were excluded from the analysis. In the Fig. l(a) are shown: (1) the maximum # (corrected geomagnetic latitude) value of closed field lines of the geomagnetic field according to Alouette-2, (2) the border of the region of trapped radiation at morning hours according to Alouette-1, and (3) the border of the region, which is characterized by a rapid decrease of the flux of electrons with E > 35 keV in midnight hours. The position of the aurora1 oval with lir, = O-l during day and night hours according to Feldstein (1966) is marked by oblique shading. The latitude of most frequent occurrence of aurora in the zenith is marked by a dashed line. Direct comparison shows that the oval is situated near the border of closed field lines. If the latitude of rapid reduction of counting rate is taken (to l/10 in an interval 501
Y. I. FELDSTEIN
and G. V. STARKOV
l(a). THIZCGM~ARLWN~PTHEPOSITION~FTHEAURORAL~~ALWITHTHEDATA~F~~WWC~ SATELLITE%
(1) The limit of closed geomagnetic field lines according to McDiarmid et 01. (1968a). (2) The boundary of the region of trapped radiation in morning hours according to McDiarmid et al. (1964). (3) The boundary of a rapid decrease of the flux of electrons with E > 35 keV in midnight hours according to McDiarmid et al. (1968a). Oblique shading marks the amoral oval (P > 60 per cent) with K, = O-l in day and night periods according to Feldstein (1966). Straight shading marks the aurora1 oval with Q = 0.
LGT FIG. l(b). THE STRINGER
et al. (1967) AND
(SHADED) IN THE PERIOD DURING
THE LIMIT OF CLGSED GEOMAGNETIC AL0Urrrra-2 SAT6LLTl-B.
PIELD LINES ACCORDING
Curves1 and 3 are the same as in Fig. l(a).
LGT FIC3. l(C).
(MCDIARMID et al. 1968a). Curves 1 and 3 are the same as in Fig. 1 (a). The dotted lines show the position of the auroral orientation curves according to Gustafsson (1967) situated on 4’ = 69” and +’ = 70-S” in midnight.
of less than 0.5”) as the border of the region of trapped radiation, then the oval is situated just on the poleward side of the border of the trapped radiation region of electrons with E > 35 keV. This is practically the border of the closed field lines of the geomagnetic field during the whole day. In Fig. l(a) the position of the border of the oval according to Feldstein et al. (1967) with Q = 0 is marked by straight shading. The amoral oval is situated on the border of the region of closed field lines during the whole day. Only during morning hours the oval border is somewhat closer to the equator than in other local hours. It is worth noting that in morning hours the border of the trappingregion ofelectrons of energies E > 40 keV, obtained by the Alouette-1 (October 1962-March 1963), is also closer to the equator. Taking into account the fact that r.m.s. deviations in determining the border of the aurora1 oval is of the order of l-1.5” it is suggested that the equatorward boundary of the aurora1 oval lies close to the boundary of closed field lines. The position of the boundary of closed field lines was obtained by the Alouette-2 satellite during the period of the nearly minimum solar activity. Therefore, more correct results can also be obtained by comparing the position of the aurora1 oval with the trapping boundary during the period of the minimum solar activity. In Fig. l(b), the results of this comparison are shown. The boundary of closed field lines is taken from the work by McDiarmid et al. (1968a), while the oval from the work by Stringer et al. (1967) for the IQSY period. In such a case the isolines of a relative number of auroral forms in lo interval of geomagnetic latitude are taken for aurora1 zone boundaries. The maximum number of discrete auroral forms does really appear near the boundary of closed force-lines. Gustafsson (1967), Feldstein et al. (1967) obtained the position of auroral orientation curves using observations of variations of the azimuth of arcs and bands in different latitudes during the course of a day. In Fig. l(c) the position of auroral orientation curves which are situated in zenith on $’ = 69” and $’ = 70.5” after Gustafsson (1967) during midnight
I. FELDSTEIN and G. V. STARKOV
hours is marked by a dotted line. If those two aurora1 orientation curves are situated at the boundary of closed field lines at nearmidnight hours, they will be situated outside the trapped region at all local hours. The equatorial border of the auroral oval on the night side practically coincides with the border of the westward electrojet. If the auroral oval is situated outside the region of trapped electrons with E > 40 keV, then the electrojet must be situated near the poleward boundary of the trapped region. According to Craven (1967), the high latitude boundary of trapped electrons with E > 40 keV at the height of 1500 km (according to satellite Injun-3) is displaced equatorwards during bay-like magnetic disturbances. This displacement was observed at all local hours, and the largest displacement took place on the night side of the Earth. The equivalent line currents at the height of the ionosphere, which may be considered to be responsible for observed disturbances of geomagnetic field, were situated some degrees northward from the trapping boundary of 40 keV electrons. Afonina ef al. (1969) came to the same conclusion about the position of currents in the ionosphere for the period of negative bay disturbances, during the flight period of satellite 1963 38C (Williams et al., 1966). Therefore, one may conclude that both the auroral oval and the westward electrojet in the ionosphere are situated along the poleward side of the boundary of the closed geomagnetic field lines, i.e. on the force-lines forming the magnetospheric tail. The comparison of #* and #c, determined on the basis of measurements by satellite Alouette-2, with the position of auroras for different levels of magnetic activity was also made for the midnight and midday part of the oval. (4, coincides with the boundary of the stable trapping region, and #0 with the poleward boundary of the region, whichis determined by the invariant latitude where the counting rate in the E > 35 keV electron detector falls to the cosmic-ray background level. The values of +11and (50for particular moments in the period from December 1965 to February 1966, according to satellite Alouette-2, which McDiarmid kindly reported to us, were averaged for different values of Q-index in the night side of the Earth on f N 65”. In Fig. 2(a) the results of the comparison of (b, and di, with the position of the aurora1 oval in the midnight sector for the IQSY period, obtained by Feldstein et al. (1968a), are shown. The boundary of the stable trapping region coincides with the equatorward boundary of the auroral oval to approximately a tenth of a degree, regardless of the degree of magnetic activity within the accuracy of 1”. The values of #e are in agreement with the latitude of the poleward boundary of the oval, though less clear than the relation between +S and the equatorward boundary of the oval. Therefore, we may conclude that the corpuscular stream, which causes auroras in the midnight sector, is generated in the region of the magnetosphere which is projected along field lines to the interval of latitudes between A and #e. Moreover, the size of this particular region enlarges with the increasing of magnetic activity. This may correspond with the simultaneous widening of the area in the tail region (in the direction to and from the Earth), in which the acceleration of electrons to the energy of several keV takes place. The small difference between +Sand d in calm magnetic periods shows that the acceleration of particles takes place in a narrow region. Based on the results obtained in the above, some conclusions may be drawn on the dynamics of the large-scale structure of the magnetosphere during the development phase of magnetic disturbances. The position of the aurora1 oval during the development of polar magnetic disturbances (DP) and the ring current (L)R) was analyzed by Feldstein ef al.
2(a). Tm3 CHANGING
OF MAGNBTIC TO
(1968a) 35keV--THB al.
ZENITH FOR +,
FIG. 2(b). THE SAME FOR MIDDAY HOURS. The aurora1 oval for the IGY period according to Feldstein et al. (1968b). The values of d in 0200-1000 br UT are marked by triangles, in 1500-2000 hr UT by light circles. Black circles is value of +#, recounted to one value of angle 6.
(1967, 1968). As it can be presumed that 4, coincides with the equatorial boundary of the aurora1 oval during midnight hours, then with an intensification of DP a reduction of the volume of space of the dipole character on the night side of the Earth occurred. Simultaneously, the region in which the streams of electrons with the energy of several keV are generated widens. During the development phase of the ring current only the approaching of the boundaries of the closed force-lines region to the Earth is observed. The dimension of the region where the electron acceleration occurs does not change. On the basis of the Injun-3 satellite, Maehlum (1968) compared the streams of trapped and precipitated electrons with E > 40 keV with the intensity of emission i13914A fer
Y. I. FELDSTEIN and G. V. STARKOV
quiet magnetic periods. During near-midnight hours, a sharp poleward boundary of the trapped and precipitated electrons was always observed; moreover, the precipitation took place in a narrow region near this border. The luminosity of emission 13914 A was seen near the poleward boundary of the stable trapping. The results of comparison for the midday part of the oval are shown in Fig. 2. The position of the auroral belt is taken for the period of IGY (Feldstein et al., 1968), but without a significant DR. Since the difference between the position of the aurora1 oval during IGY and IQSY with the same intensity of DP may be caused by the difference of DR (Feldstein et al., 1968a) for the two periods, it can be presumed that the position of the aurora1 belt Fig. 2 corresponds with the IQSY period rather well. According to data which we have in our disposal, the passing of a satellite above the midday part of the auroral oval took place mainly in 0200-1000 hr and 1500-2000 hr UT. It was observed that for the same intensity of DP the difference in the values of & for different intervals of universal time was the order of 3”. In Fig. 2 the values of I& for the period of 0200-1000 hr UT are maked by triangles, and for the period of 1500-2000 hr UT by circles. This difference shows the existence of a longitude effect in the position of the boundary of closed field lines in nearmidday hours. Average values of & for eight 3-hr intervals of UT for quiet geomagnetic conditions (Q < 3) for th e intervals 1000-1400 hr and 1400-1800 hr LT were calculated to determine the magnitude of the UT dependence. According to Feldstein et al. (1967) the position of the aurora1 belt in these two intervals of local time with such values of Q is practically the same. The obtained results are shown on the Fig. 3(a). The curve of change of $’ as a function of UT is approximately sinusoidal. The maximum values of $’ are observed in 15001800 hr, the minimum in 0300-0600 hr UT. A similar longitude effect in the intensity of magnetic activity at C#N 67” was obtained by Feldstein (1963). The variation of the position of the boundary of the precipitation of the low energy electron region as a function of UT was discovered by Maehlum (1968); his study was based on the Aurora-l satellite data. For 4’ N 75” on the day and night part of the Earth in the northern hemisphere, the maximum of low energy electron stream is observed in the 1800-2000 hr UT sector and the minimum in the 0600-1200 hr UT sector (Maehlum, 1968). It was assumed that this effect is caused by the daily variation of the orientation of the geomagnetic field with respect to the ecliptic plane. In that case a relationship should exist between +C on the day-side of the oval zone and the &angle between the Sun-Earth direction and the geomagnetic dipole axis. The angle 6 is ‘the function of UT and the season. Average values of +Cfor 5” intervals of angle 6 are shown in Fig. 3(b). The dependence of & on 8 is approximately given by the form of C&= 79.2” - 0,15(90” - 8). This is shown in Fig. 3(b) by a dashed line. Some difference between the results, obtained by McDiarmid et sl. (1968) and ours may be caused by the fact that the averaging of +Cin the study by McDiarmid et al. (1968b) was done for the period 0200-1400 hr LT in a large range of changes of magnetic disturbance intensity (up to K,, = 4,). The position of the midday part of the auroral belt was obtained by Feldstein ef al. (1967); their data were based on records from Pyramida in December and January of 1957-59. For this particular set of conditions, the angle between the dipole axis and ecliptic plane is comparatively stable. The angle 6 was 66-70”. For this reason all the
FIG. 3(a). THE CHANGINGOF +D WITH UT IN locK%18~hr LT DURING THE QUIET MAGNETIC PIELD. FIG. 3(b). TDP DEPENDENCE OF I#~M DAY HOURSFROMTHE ANGLE 6 BETWEENTHE DIRJKX’ION OF SUN-EARTH AND GEOMAGNETTC DIPOLEAXIS.
values of & in midday hours according to Alouette-2, were recauculated in accordance with marked correlation to 6 = 68”. The obtained average values of & for 6 = 68’ with different Q indexes are shown in Fig. 2(b) by black circles and the r.m.s. deviation is marked by straight lines. For each Q index the value of I$~is situated within the aurora1 belt. On the basis of the results obtained in the above, the equatorward boundary of the aurora1 oval may be identified with the position of the poleward boundary of the region of closed geomagnetic field lines. The longitude effect, which consists in the change of the latitude of the boundary on the day side, is caused by the change of the orientation of the dipole axis in the streaming corpuscular stream. Acknowledgements-We wish to thank Dr. I. B. McDiarmid for permission to use his data from the Alouette-2 satellite and the referee for the improvement of the translation from Russian into English. REFERENCES APONINA,R. G. and FELDSTEIN,Y. I. (1969). The boundary of trapped E > 280 keV electrons and electrojets location in the ionosphere. Space Invest. 7, 312-314. AKAWPD, S. I. (1966). The aurora1 oval, the auroral substorm and their relations with the internal structure of the magnetosphere. Planet. Space Sci. 14,587-59X CRAVEN, J. D. (1967). Low-altitude observations of the high-latitude cutoff of electron (E > 40 keV) intensities during magnetic bays in the aurora1 zones. Trans. Am. geophys. Un. 48, 181. FELDSTEIN,Y. I. (1963). Space-time distribution of magnetic activity at high latitudes of the Northern Hemisphere-III. ZGY Program, Vol. 5, pp. S-63. Acad. Sci. U.S.S.R., Moscow. FELDSTEIN,Y. I. (1966). Some peculiarities in aurora and magnetic disturbances distribution in high latitudes, caused by the asymmetrical form of the magnetosphere. Planet. Spuce Sci. 14,121-130. FELDSTEIN,Y. I. (1967). Distribution of aurora and magnetic disturbances at high latitude in connection with the asymmetric form of the magnetosphere. Aurora and Airglow, Vol. 13, pp. 98-118. Nauka, Moscow. FELDSTEIN,Y. I. and STARKOV,G. V. (1967). Dynamics of aurora1 belt and polar geomagnetic disturbances. Pldnet. Space Sci. 15,209-229. FELDSTP~N,Y. I. and STARKOV,G. V. (1968a). Auroral oval in the IGY and IQSY period and a ring current in the magetosphere. Planet. Space Sci. 16,129133. FELDSTEIN,Y. I., SHEVNIN,A. D. and STARKOV,G. V. (1968b). Auroral oval and magnetic field in the tail of the magnetosphere. Annls Gdophys. 24,517-520. GUSTAFSSON,G. (1967). On the orientation of aurora1 arcs. Plunet. Space Sci. 15,277-294.
Y. I. FELDSTEIN
and G. V. STARKOV
~HLXJM, B. N. (1968a). Satellite and rocket observationsin the auroralzone. Annls Gbophys. 24,275-284. MAEMUM, B. N. (1968b). Universal time control of the low energy electron fluxes in the polar regions. J. geophys. Res. 73,345%3468. MCDIARMID, I. B. and Bunaows, J. R. (1964). High latitude boundary of the outer radiation zone at 1000 km. Can. J. Phys. 42,616-626. MCDIARMID,I. B. and BARROWS,J. R. (1968a). Local time asymmetries in the high latitude boundary of the outer radiation zone for different electron energies. Can. J. Phys. 46,49-62. McDm, I. B. and WILSON, M. D. (1968b). Dependence of the high latitude electron (E > 35 kev) boundary on the orientation of the geomagnetic axis. J. geophys. Res. 73,7237-7244. PIDDINGTON, J. H. (1965). The magnetosphere and its envirous. Planet. space Sci. 13, 363-376. STRINGER,W. J. and BELON,A. E. (1967). The morphology of the IQSY auroral oval. I. Interpretation of isoauroral diagrams. J. geophys. Res. 72,4415-4421. WILL-, D. J. and NESS, N. F. (1966). Simultaneous trapped electron and magnetic tail field observations. J. geophys. Res. 71,5117-5128.