We examined the effect of magnetic disturbances in two polar caps on the generation of magnetospheric substorms. For this purpose we investigated the correlation between the AL index (showing substorm activity in the Northern hemisphere) and two geomagnetic activity indices, the Polar Cap (PC) index and Polar Magnetic (PM) index showing the magnetic disturbances in the Northern and Southern polar caps. For the analysis we used the data for four years when geomagnetic activity indices were available in both hemispheres. We obtained an unexpected yet important result: while in northern winter the correlation between AL index and northern PC/PM indices is very good, in northern summer the AL index correlates much better with southern PC/PM indices. Thus, substorm activity in summer months correlates much better with geomagnetic activity not in the nearby polar cap but in the opposite polar cap. This effect may be caused by the interhemispheric field‐aligned currents flowing from the summer high‐latitude ionosphere and closing through the ionosphere in the opposite auroral zone. An interesting feature of these interhemispheric currents is that they are directed opposite to the substorm field‐aligned currents in the summer hemisphere but along the substorm field‐aligned currents in the winter hemisphere. This leads to decreasing the total field‐aligned currents and their contribution to magnetic disturbances in the summer hemisphere but increasing these currents and related magnetic disturbances in winter hemisphere, which explains the experimental results obtained in our study.

[1] From a simple theoretical consideration, we obtained two coupling functions linking upstream solar wind parameters to geomagnetic activity. We took into account (1) a scaling factor related to polar cap expansion while increasing the reconnected magnetic flux in the dayside magnetosphere, and (2) a modified Akasofu function for the reconnected flux for combined IMF B z and B y components. One of these coupling functions may be written as F a = aV sw B yz 1/2 sin a (q/2), where V sw is the solar wind speed, B yz is the magnitude of the IMF vector in the Y-Z plane, q is the clock angle between the Z axis and IMF vector in the Y-Z plane, a is a coefficient, and the exponent, a, is derived from the experimental data and equals approximately to 2. The F a function is proportional to the square root of B yz , which makes it significantly different from the coupling functions proposed earlier. Nevertheless, the statistical data analysis supports this dependence. For testing the found coupling function, we used the solar wind and IMF data for four years. We computed 2-D diagrams showing the correlation coefficients for the dependence of the polar cap PC geomagnetic activity index on different combinations of solar wind/IMF parameters. The obtained diagrams showed very good agreement with the theoretical coupling function. The correlation coefficient for the dependence of the PC index on the coupling function is about 0.8-0.85, which is significantly higher that that for other commonly used coupling functions for the same time intervals.

[1] Earlier we found that the auroral electrojet AL index, indicating substorm activity in the northern hemisphere, in local summer months correlates much better with geomagnetic activity not in the nearby polar cap but in the opposite polar cap; we explained this effect as a result of interhemispheric field-aligned currents, which suppress substorm field-aligned currents in the summer hemisphere but increase these currents in the winter hemisphere. In the present paper, we took into account this effect and examined a method for reliably monitoring the substorm auroral electrojet, measured with the auroral electrojet AL index, by using hourly averages of geomagnetic field from two polar observatories (Thule and Vostok) in two hemispheres. We tested this method for 3 years. The correlation between the predicted and actual AL indices for these years was stable and very high, and it showed no significant dependence on season and a relatively weak UT variation. The correlation coefficient between the predicted and actual AL indices for these three years was about 0.89. The proposed method, based on using magnetic field data from two polar geomagnetic observatories in two hemispheres, not only significantly improves the reliability of monitoring the westward auroral electrojet in the northern hemisphere but it may also be used for monitoring the westward auroral electrojet in the southern hemisphere where no AL index is available because a significant portion of the southern auroral zone is located over the oceans. The results of this paper show that measurements of geomagnetic field in two hemispheres are of high importance for reliably monitoring the geomagnetic activity and related events in each hemisphere.

An asymmetry in ionospheric conductivity between two hemispheres results in the formation of additional, interhemispheric field-aligned currents (FACs) flowing between conjugate ionospheres within two auroral zones. These interhemispheric currents are especially significant during summer-winter conditions when there is a significant asymmetry in ionospheric conductivity in two hemispheres. In such conditions, these currents may be comparable in magnitude with the Region 1 (R1) field-aligned currents. In this case, the R1 current is the sum of two FACs: one is going from/to the solar wind, and another is flowing between conjugate ionospheres. These interhemispheric currents can also cause the formation of auroras extended along the nightside polar cap boundary, which may be related to the so-called "double auroral oval." In this study, we present the results of analytical and numerical solutions for the interhemispheric currents and their effect on the Region 1 currents.