Because of the solar wind originating from the Sun, Earth is hit by a hot, magnetized, supersonic collisionless plasma carrying a large amount of kinetic and electrical energy. Some of this energy finds its way into our magnetosphere creating, e.g., geomagnetic activity which consists of
When geomagnetic activity has any practical importance to human technology etc., we talk about space weather.
The storms are most directly related to specific solar wind events, while the substorm activity is more complicated because of the temporal storing of energy in the magnetotail. It is not necessary to have a storm in order to have a substorm! Note also that while some auroras (those that extend to low latitudes) are storm-time features and some others (the most active ones) relate to substorms, the oval does not disappear even during the more quiet magnetospheric periods.
It has been shown that solar wind speed correlates well with geomagnetic activity at time scales longer than about one month (Gosling et al., 1976; Crooker et al., 1977). Also the IMF affects the geomagnetic activity, although the energy density of the magnetic field is small in comparison with that of the solar wind plasma. This is because the southward IMF component enhances the coupling between the solar wind and the magnetosphere/ionosphere system.
The level of the geomagnetic activity is measured using different activity indices, most of which are based on ground-based magnetometer recordings. These recordings can be used, e.g., to study the longer trends in the solar activity (e.g., Russell, 1975). Variability in the geomagnetic activity has several sources:
These effects will be discussed more below.
The 22-year double-solar-cycle variation in geomagnetic activity was identified by Chernosky (1966). Activity is higher in the second half of even-numbered solar cycles and in the first half of odd-numbered cycles. The reasons for this is still under discussion: Cliver et al. (1996) argue that it is intrinsic solar variation not related to Russell-McPherron or Rosenberg-Coleman polarity effect as typically suggested (see semi-annual variability below).
The 11-year variability of the geomagnetic activity (e.g., Ellis, 1900) has been recently studied by Vennestrom and Friis-Christensen (1996). They suggest that the activity can be divided into three peaks:
See how the two peaks, one somewhat ahead or at solar maximum and the other 2 or 3 years after it, can be seen in the SSC frequency.
The 1.3-1.4-year variability originating from the Sun has been observed in the geomagnetic or auroral data by, e.g., Shapiro (1967), Silverman and Shapiro (1983), and Paularena et al. (1995).
The annual geomagnetic variation relates to the Earth's orbit. Due to the 7.2 degrees tilt of the solar rotation axis with respect to the normal of ecliptic, the Earth reaches the highest northern and southern heliographic latitude (where solar wind speed is higher) on September 6 and March 5, respectively, and crosses the equator twice a year between these dates. Thus, when observed from Earth, one should expect a semiannual variation in solar wind speed with maxima around these dates. However, annual variation is often more clear (e.g., Bolton, 1990), and this is because the solar wind distribution is asymmetric or shifted with respect to equator (Zieger and Mursula, 1998).
The semi-annual variation has been attributed to a IMF-effect (Russell-McPherron, 1973): as the Earth orbits around the Sun, southward IMF component is statistically more likely twice a year, increasing the coupling between the solar wind and magnetosphere. As a result, more storms occur during equinoctial months than during the solstitial months.
The relationship between geomagnetic/auroral activity and solar rotation period of 27 days was noted well before space age (e.g., Broun, 1876; Maunder, 1905). This recurrent storm activity is due to coronal holes that cause fast solar wind streams (term M-region was used earlier; see also Crooker and Cliver, 1994). In addition, long intervals exists when two high-speed streams per solar rotation can be seen (e.g., Gosling et al., 1976), creating a 13.5-day periodicity. See, e.g., Mursula and Zieger (1996,1998) for more discussion about the matter.