Aurora is now known to be caused by electrons of typical energy of 1-15 keV, i.e. the energy obtained by the electrons passing through a voltage difference of 1000-15,000 volts. The light is produced when they collide with atoms of the upper atmosphere, typically at altitudes of 80-150 km. It tends to be dominated by emissions of atomic oxygen--the greenish line at 5577 A and (especially with electrons of lower energy and higher altitude) the dark-red line at 6300 A. Both these represent "forbidden" transitions of atomic oxygen from energy levels which (in absence of collisions) persist for a long time, accounting for the slow brightening and fading (0.5-1 sec) of auroral rays. Many other lines can also be observed, especially those of molecular nitrogen, and these vary much faster, revealing the true dynamic nature of the aurora.
Aurora can also be observed in the ultra-violet (UV) light, a very good way of observing it from space (but not from ground--the atmosphere absorbs UV). The "Polar" spacecraft even observed it in X-rays. The image is very rough, but precipitation of high-energy electrons can be identified.
The curtains often show folds called "striations." When the field line guiding a bright auroral patch leads to a point directly above the observer, the aurora may appear as a "corona" of diverging rays, an effect of perspective.
In 1741 Hiorter and Celsius first noticed other evidence for magnetic control, namely, large magnetic fluctuations occurred whenever the aurora was observed overhead. This indicates (it was later realized) that large electric
currents were associated with the aurora, flowing in the region where auroral light originated. Kristian Birkeland (1903) deduced that the currents flowed in the east-west directions along the auroral arc
, and such currents, flowing from the dayside towards (approximately) midnight were later named "auroral electrojets." (see also Birkeland currents).
Still another evidence for a magnetic connection are the statistics of auroral observations. Elias Loomis (1860) and later in more detail Hermann Fritz (1881) established that aurora appeared mainly in the "auroral zone," a ring-shaped region of approx. radius 2500 km around the magnetic pole of the Earth, not its geographic one. It was hardly ever seen near that pole itself. The instantaneous distribution of aurora ("auroral oval," Yasha Feldstein 1963) is slightly different, centered about 3-5 degrees nightward of the magnetic pole, so that auroral arcs reach furthest equatorward around midnight.
Aurora is a common occurrence in the ring-shaped zone. it is rarely seen in temperate latitudes, usually only when a big magnetic storm temporarily expands the auroral oval. Large magnetic storms are most common during the peak of the 11-year sunspot cycle, or during the 3 years after that peak. However, within the auroral zone the likelihood of aurora depends mostly on the slant of interplanetary magnetic field lines (further below), being greater with southward slant.