A more frequent type of disturbance, and perhaps a more fundamental one, is the magnetic substorm (its relation to magnetic storms is discussed further below). Many features of the substorm remain to be explained, but it is widely held to represent the violent discharge of magnetic energy, accumulated in the tail of the magnetosphere.
Substorms were recognized much later than magnetic storms, because their effects are observed on the ground mainly in the auroral zone. After observing a great auroral display in Connecticut on July 1, 1837, E.C. Herrick wrote (italics in the original):
"It is worthy of notice, that on this occasion there were two well-marked seasons of greatest brilliance or fits of maximum intensity, at intervals of about four hours. It will be found on examination of former accounts, that this is a common feature of Auroral exhibitions of unusual brilliance."
As the Norwegian scientist Kristian Birkeland observed around 1900, those "fits of maximum intensity" also created magnetic disturbances on the ground. After analyzing them with a network of four stations, he concluded that they came from large electric currents flowing along the auroral arcs. Birkeland regarded them as a new phenomenon, which he named an "elementary polar magnetic storm" and which typically lasted half an hour.
In 1964 S.-I. Akasofu named them "magnetic substorms," a term originally introduced by his mentor Sidney Chapman, who viewed them as phases of magnetic storms. Akasofu noted that they often occured independently of storms and that tended to follow a certain pattern--the aurora brightened, its arcs moved rapidly and spread across the sky (especially polewards), then broke up and dwindled away. The associated magnetic disturbance could be quite large, maybe 3-10 times that of magnetic storms, but it dropped down quickly outside the auroral zone, implying that it was produced by nearby electric currents, as Birkeland had also concluded.
Satellites in space see much more profound changes than the ground observer. The number of particles observed on the night side, in synchronous orbit and in the tail, may increase a hundred-fold and more, or else the particles may disappear altogether, suggesting a drastic re-shaping of magnetic field lines. Images from satellite cameras show a band of active and bright aurora over the night side of the polar cap, low-altitude satellites passing that band intercept intense flows of electrons, and satellites in the plasma sheet detect fast and abrupt ion flows.
All this suggests that the substorm accelerates ions and electrons to higher energies, and for this it has received more attention than any other magnetospheric phenomenon. Scientists believe that like a flare on the Sun, the substorm represents the sudden conversion of magnetic energy to particle energy, and in fact the name "auroral flare" was once suggested. Many also believe it to be associated with magnetic reconnection, a process originally proposed for flares but later applied to the Earth's magnetopause. Yet much is still uncertain, in part because most data on substorms in space come from isolated spacecraft which cannot see the full pattern, and several rival theories of the substorm still exist.
Electric Currents from Space
They represent a new and altogether different type of electric current in space. Other currents described so far--the ring current, the current on the magnetopause which confines the Earth's field, and the plasma sheet current which flows across the tail--all are "coasting currents" and (in principle) consume no energy. As long as a plasma is present in a suitable magnetic field, the spiraling and drifting motions of its ions and electrons automatically produce the electric current flow.
Electric currents which flow in and out of the ionosphere are more like house currents, which need a voltage (a kind of electrical pressure) to drive them through wires and appliances, and which do transmit energy. Indeed, Birkeland currents might be one of the avenues by which the magnetosphere extracts energy from the solar wind. It is time for a closer look at the processes involved.
Picture of Faraday's disk dynamo.
"Dynamo processes" can also exist when the circuit is formed by a plasma, part of which flows across a magnetic field, e.g. plasma of the ionosphere or the magnetosphere. In particular, if some of the Earth's magnetic field lines are "open" and extend into the solar wind, a dynamo circuit should exist: it would include (picture below) a path in the ionosphere between two open field lines, the lines themselves, and a closure path in the solar wind. Since the solar wind moves relative to the ionosphere (or vice versa), the conditions for a dynamo exist and a current is expected to flow; as part of the process, the solar wind in the distant leg of the circuit is slightly slowed down, and this supplies the energy needed to produce the current. Such a "dynamo" may in fact explain at least some of the Birkeland currents.
Cartoon of dynamo involving open field lines.
The acceleration of these new ions and their injection into the ring current is accompanied by large and violent substorms. The auroral region expands, allowing aurora to be seen equatorward from its usual location. That is the reason why many people associate auroras with magnetic storms!
We still do not know why substorms occuring in magnetic storms differ from the usual type, why they inject much more plasma into the inner magnetosphere and to a much greater depth. For one thing, "ordinary" substorms are triggered by relatively moderate changes in the solar wind or IMF, and may even erupt spontaneously if favorable conditions exist. Full-size magnetic storms always require a strong interplanetary stimulation--the arrival of a fast and dense plasma cloud, launched by a solar eruption (the cloud's magnetic field should have a southward slant), or of a fast stream of the solar wind. They are relatively infrequent, occuring perhaps once or twice a month--nothing like Herrick's four-hour interval, cited at the start of this section.