91av

On the night of 13 March 1989, strange lights appeared in the sky all
over southern England. They were part of an unusually extensive aurora borealis,
or northern lights, the colourful display normally restricted to the polar
skies. This rare appearance at such low latitudes was a sign that the Earth
had picked up more energy than usual from the solar wind. Next Friday, three
years later to the day, researchers will be meeting at the Royal Astronomical
Society in London to discuss, among other things, how and why this happened.

The meeting will consider the progress and plans of an international
research effort bent on understanding energy transfer in the Solar System,
the Solar-Terrestrial Energy Programme, known as STEP. The program will
coordinate the diverse groups of researchers whose work has a bearing on
the mechanisms by which energy reaches the Earth from the Solar wind. The
project will last at least seven years. Combining results from various satellite
missions and observatories on the ground with theory and computer simulations,
the STEP researchers hope to find out more about the behaviour of ionised
gases, called plasmas. This form of matter dominates the known cosmos, forming
more than 90 per cent of the matter we can detect.

There is plasma in stars, in space and around the Earth, including a
part of the upper atmosphere that we term the ionosphere. Plasma in interplanetary
space comes from the Sun; most of the energy produced by nuclear fusion
radiates into space as electromagnetic waves. But some is carried away from
the Sun as kinetic energy by a stream of charged particles – the solar wind.
Some of these particles and the energy they carry penetrate the Earth’s
magnetic field and trigger the displays we see as auroras.

The idea that bursts of particles emanating from the Sun are responsible
for disturbances in the Earth’s magnetic field was first suggested in 1931.
Observations of the ionised parts of comet tails – which always point away
from the Sun, as if blown that way – led in the 1950s to suggestions that
the solar wind is always present. This was later confirmed by satellite
measurements. The solar wind streams away from the Sun, driven by the very
high temperatures of plasma in the Sun’s corona, at about 1 000 000 °C.
The plasma making up the solar wind permeates a vast region around our Sun,
called the heliosphere. No one yet knows where the heliosphere ends, as
no spacecraft sending back information from outside the Solar System has
yet encountered this boundary.

The Earth’s magnetic field is an obstacle to the solar wind, and only
a fraction of the energy flowing through interplanetary space ordinarily
comes near our planet. The process that links us with the heliosphere was
first described by the British physicist J. Dungey in 1961, in a paper just
two pages long. From these two pages has grown much of our understanding
of the transfer of energy from the solar wind to the Earth. But many physicists
disagreed with Dungey’s ideas at the time and they have argued over them
since. One of the problems is the sheer scale of the phenomena; they stretch
around the world and far out into space, and most observations see only
a fraction of what is going on.

There is no shortage of puzzles in the area that STEP covers. Many of
them focus on fluctuations in the solar wind, springing from the enormous
variability of the Sun itself, which is signalled by sunspots and solar
flares. So although the solar wind is continuous, it is far from constant
in velocity or density.

Better techniques for observation are providing striking illustrations
of this variability. The sequence of images in Figure 1 shows what is known
as a coronal mass ejection, in which a huge bubble of enhanced plasma bursts
out from the Sun in just two hours. On average, such events release about
1013 kilograms of matter and at first move at about 350 kilometres per second.
These bubbles propagate through interplanetary space, like the quiet solar
wind, and can buffet the Earth. Each one carries about 5 x 10 ¹⁸
watts of power, the equivalent of five thousand million power stations.
But although they happen every two days, on average, they are not all directed
at the Earth.

Signs of similar events spreading through interplanetary space were
recorded by the Giotto spacecraft on its way to rendezvous with Halley’s
comet. Giotto’s measurements were combined with others from Soviet, Japanese
and US spacecraft as well as from observatories on the ground. STEP will
help to combine observations such as these to refine theoretical models
of the shape and motion of solar wind structures and how energy moves through
the Solar System. From these models, scientists are attempting to predict
how the solar wind varies; they want to be able to forecast the ‘interplanetary
·É±ð²¹³Ù³ó±ð°ù’.

These models often attempt to describe the behaviour of the bulk of
plasma over most of the heliosphere. Because the plasma is made up of moving
charged particles, giving electric and magnetic fields, its behaviour is
described by Maxwell’s equations, which relate electric and magnetic phenomena.
From this starting point, together with a form of Ohm’s law relating current
and electric field, researchers have evolved a theoretical description of
plasma behaviour known as magnetohydrodynamics. According to this description,
if plasma is spread over a large area, and its electrical conductivity is
high, the magnetic field and the plasma always move together – researchers
say that the magnetic field is ‘frozen-in’ to the plasma. This is what happens
in the Sun’s corona and in interplanetary space, for example.

A weak magnetic field originating from the Sun permeates the entire
heliosphere; it is called the interplanetary magnetic field. The usual way
to visualise a magnetic field is as a pattern of lines that run as closed
loops from the south to the north pole of the source of the field (as if
for a bar magnet). Around the Earth, these loops link the northern and southern
hemispheres; in interplanetary space they run to and from the Sun. Over
most of the heliosphere, the density of kinetic energy in the solar wind
is much higher than the density of energy stored in the interplanetary field,
with the result that the plasma dominates. Because the field lines are frozen
in, the flow of the plasma drags them away from the Sun, like huge rubber
bands.

Some planets, such as Earth, have magnetic fields that store a higher
density of energy than the plasma nearby, so the situation is reversed.
Here plasma follows the field lines and is trapped around the planet in
a region known as the magnetosphere, shown for Earth in Figure 2.

This connection between plasma and magnetic fields is a simplification
of solar-planetary physics and does not apply everywhere. It breaks down
in explosive solar phenomena such as flares and at the edges of planetary
magnetospheres. The most important process in the breakdown is known as
‘magnetic reconnection’.

Reconnection takes place on a large scale about 50 000 kilometres from
the Earth, where the flow of the solar wind meets the edge of the Earth’s
magnetosphere, the ‘magneto-pause’. Because the solar wind is supersonic,
a shock front forms upwind of the magnetopause. The flow slows behind this
‘bow shock’ but carries on unabated away from the Earth. The result is that
the lines of the interplanetary field, frozen into the solar wind, become
draped over the nose of the magnetosphere. The magnetosphere is compressed
so tightly that the frozen-in condition breaks down locally, and with it
the barrier that the magnetopause presents to the solar wind. This transfers
mass, momentum and energy across the boundary from the solar wind into the
magnetosphere.

In order to understand reconnection, it is important to realise that
although magnetic field lines cannot be cut, two loops can fuse together
and a new configuration emerge. Reconnection can take two small loops and
generate one large one, via a figure-of-eight shape, or it can do the reverse.
But reconnection can only happen where two field lines, in opposing directions,
are pressed together.

The interplanetary field varies rapidly and widely in orientation with
respect to the Earth’s magnetic axis; only when it has a southward component
can reconnection take place at the nose of the magnetosphere, for there
the Earth’s field always points northward. The Earth’s field lines can then
reconfigure to produce ‘open’ field lines which cross the magnetopause and
connect the Earth with interplanetary space and the Sun.

The field lines that cross the magnetopause join interplanetary space
where the field lines are frozen into the flow of the solar wind and dragged
away from the Sun. This drag transfers momentum and energy into the magnetosphere
and solar wind particles move down the open field lines into the ionosphere,
colliding with atoms and molecules in the atmosphere and generating auroras.
And the plasma in the ionosphere is frozen with the moving field lines,
generating heat as the particles collide with neutral atmospheric particles.
Not all the energy is deposited in the atmosphere in this direct way: much
of it is stored temporarily as magnetic energy in the tail of the magnetosphere
formed as the flow of the solar wind drags open field lines away from the
Sun.

Without reconnection, the Earth’s magnetosphere would gain little energy
from the solar wind and contain only plasma produced by the action of ultraviolet
and X-rays from the Sun on the upper atmosphere. This is, in fact, the case
only for a relatively small, doughnut-shaped region close to the Earth called
the plasmasphere. The majority of the magnetosphere contains much plasma
that originated at the Sun, circulates on a large scale, and extracts huge
amounts of energy from the solar wind.

The energy carried by the solar wind is highly variable but averages
about 0.5 milliwatts per square metre. This meets the Earth’s magnetosphere,
an obstacle with an area of about 3 x 10 ¹⁶ square metres, and
supplies to the Earth, on average, a total power of about 1.5 x 10 ¹³
watts. This is ten thousand times less energy than sunlight supplies to
the Earth, but is still significant. To put it in perspective, it is delivering
one and a half times as much power as the human race is currently using
from whatever source.

Particles from the solar wind flowing along the open field lines directly
into the atmosphere make oxygen atoms emit the characteristic red light
of the aurora. Other particles enter through the tail of the magnetosphere
and are accelerated Earthward by energy from the solar wind. They produce
green and ultraviolet auroras.

Field lines that cross the magnetopause pass through the ionosphere
in two regions called the polar caps; these lie within rings around each
pole called the auroral ovals shown in Figure 3. When the interplanetary
field runs southwards, both the polar caps and the auroral ovals usually
expand and contract in cycles lasting a few hours. When they are especially
large, aurora can be seen at unusually low latitudes, as happened in March
1989.

The magnetic field close to the Earth, in the ionosphere is so large
that the currents associated with the aurora do not change it significantly.
When reconnection begins, more field lines cross the magnetopause and the
polar cap expands. This growth is equivalent to a voltage. Such a voltage
has been measured from satellites at, typically, 100 kilovolts.

The open field lines are swept into the tail by the flow of the solar
wind, increasing the density of energy stored there. At the typical voltage
of 100 kilovolts, this process stores energy in the tail at a rate of about
10 ¹² watts. This is 10 per cent of the total power of the solar
wind meeting the magnetosphere.

In addition to storing energy in the tail of the magnetosphere, reconnection
at the magnetopause drives electric currents in the ionosphere. The total
current flowing can be estimated at 5 million amps, which, across a potential
difference of 100 kilovolts, dissipates 5 x 10 ¹¹ watts.

At these rates, the tail of the magnetosphere can accumulate about 3
x 10 ¹⁵ joules in an hour, but not indefinitely. The polar caps
often expand, but always retreat back towards the poles. This is because
reconnection also happens in the centre of the tail, converting pairs of
open field lines back into closed field lines within the magnetosphere and
releasing the stored energy. Sometimes this happens in several places, to
form loops of field lines. These are called ‘plasmoids’; they are ejected
from the Earth down the far reaches of the tail. The trigger for the burst
of tail reconnection is not understood; but if the polar cap is to shrink,
the energy must be released faster than it accumulated. This happens in
a half-hour burst called a substorm expansion phase. Some energy is deposited
in the plasma sheet particles and gives enhanced auroral displays. The rest
generates heat and boosts circulation in the ionosphere.

The extraction of solar wind energy described here is all made possible
by reconnection at the nose of the magnetosphere on the day side of Earth.
According to theory, such reconnection should not happen when the interplanetary
magnetic field points northward, so no energy transfer should then take
place. In practice, a northward interplanetary field can reconnect with
field lines that are already open in the tail. This stirs the polar cap
plasma, and deposits some energy directly at very high latitudes. But because
this generates no extra open field lines, no energy is stored in the tail.
The overall theory is vindicated by observations such as those taken in
1984 and 1985 using the European Incoherent Scatter radars (EISCAT), in
northern Scandinavia, with the Active Magnetospheric Particle Tracer Explorer
satellites.

Understanding reconnection is just one of the goals of STEP. The distortion
of the Earth’s magnetic field also gives researchers a chance to study the
dynamics of vast regions of the magnetosphere by looking at relatively small
areas of the ionosphere around our poles. Satellites in the ionosphere yield
the most precise and objective information about the flows of plasma and
energy that result from reconnection, but they have one significant drawback.
Because they are always moving, any change they measure may result from
a change in the flow through time or a variation in space – or, as is the
usual case, both. And because lone satellites take 90 minutes or more to
orbit the Earth, they cannot detect worldwide fluctuations that happen more
quickly than this.

In contrast, groundbased observatories are well suited to studying such
variations. There has been much recent interest in using sophisticated systems
such as the EISCAT radars in Northern Scandinavia (‘Europe unscrambles the
ionosphere’, 91av, 5 December 1985), to study the rapid variations
that reveal changes in the rate of reconnection. Another approach is to
monitor the resulting heating of the ionosphere and other effects, including
the aurora, using instruments at ground level.

One of the biggest problems is exactly how reconnection links the interplanetary
magnetic field and the Earth’s field. It seems this reconnection may happen
in bursts, called flux transfer events. This possibility has been discussed
for over a decade now, particularly in the light of data from satellites
travelling close to the magnetopause on the Sunward side of the Earth. Measurements
often reveal what appear to be bumps or bulges in the magnetosphere. These
are consistent with local reconnection, resulting in bundles of recently-opened
field lines moving across the magnetopause. The best way to find out the
significance of this type of reconnection is to assess the power that it
transfers across the boundary, which depends on the size of the bumps. Unfortunately,
the satellite measurements cannot reveal the full extent of the bulges.

To find out how much energy transfer these bumps in the magnetopause
represent, the researchers had to take a different tack. The theory of reconnection
predicts that flux transfer events should produce particular patterns of
plasma flow in the ionosphere. Until recently, a vigorous search for likely
ionospheric signatures produced no strong candidates. This was a source
of some embarrassment to advocates of the standard theory that interpreted
the bumps on the magnetopause in terms of bursts of reconnection. Keen to
make the theory work, researchers postulated that these transfers may be
smaller than they had expected, giving signatures which could not be resolved
by ground-based instruments. Recently, strong evidence for flux transfer
event signatures has come from the EISCAT radars and optical observations
of the aurora on the day side of the Earth. It looks as if they evaded detection
before because, far from being small, they are considerably larger than
the field of view of most instruments.

These signatures happen only when the interplanetary magnetic field
at the Earth points southward and they recur, on average, every 7 minutes.
This is the same pattern as the behaviour of the bumps on the magnetopause.
Other evidence that suggests that these events represent significant energy
flow into the magnetosphere comes from scientists at the University of Oslo.
They have found evidence for transient events that they call ‘dayside auroral
breakup’, which also repeat on this timescale. These are difficult to see;
direct or scattered sunlight hides them, so the observations can only be
made in the dark of the polar winter. For the northern hemisphere, that
means around the December solstice from the island of Spitsbergen. The false-colour
images in Figure 4 show the speed and scale of one of these events, moving
eastward and northward across the sky. They are based on the intensity of
emissions of the red auroral light. These emissions probably come from solar
wind particles moving down newly opened field lines; the variation in the
pattern reflects changes in the rate at which field lines were reconnected.
The EISCAT radar, monitoring at the same time, showed that ion flows and
heating were dramatically increased as each feature passed, signs of the
transient energy and momentum transfer associated with each event.

This information about the extent of these flux transfer events is enough
for researchers to estimate the energy that each transmits. Assuming that
the light was emitted at an altitude of 250 kilometres, we can estimate
how many field lines reconnect in each cycle of 7 minutes, on average. The
result is that these flux transfer events contribute about 50 kilovolts
– within range of the measured value of 100 kilovolts, typical of the total
energy transfer. Theory and measurement may come even closer together in
the future, for the emission altitude, a key part of the calculation, may
have been underestimated. STEP researchers are in the process of measuring
this altitude precisely, using observations from several instruments at
different places on the ground. An increase in this altitude by a factor
of 1.5 would be enough for flux transfer events to explain all the voltage
and the energy transfer. This would be a satisfactory result for the theorists,
answering the question of how the energy passes into the magnetosphere.
But it raises a new series of questions about why reconnection shows such
quasi-periodic behaviour.

Although originally discovered by solar-terrestrial research, reconnection
is important in many other fields of science. It allows the birth of stars,
powers the explosive release of energy in flares on the surfaces of stars
and heats the Sun’s corona. In laboratories, reconnection can disrupt the
magnetic confinement of plasma and cause ‘magnetic islands’ in fusion reactors.
It may also play a crucial role in the magnetospheres of pulsars, cometary
tails and reversals of the polarity of the magnetic fields of the Sun and
the Earth.

Recent progress in the field has been dramatic. The understanding of
energy transfer in bursts of reconnection is just one of many recent discoveries,
most of which have raised more questions than they have solved. Particularly
controversial are reports of energy held in the solar wind influencing the
lower atmosphere. STEP will play a part in coordinating the operation of
satellite missions and observatories that are planned for the next decade.
Europe’s major contributions will be SOHO, the solar observatory, the CLUSTER
mission, sending four spacecraft simultaneously to the magnetosphere, and
a sister radar for EISCAT on the island of Spitsbergen. Scientists from
Britain will be involved in other missions to explore different parts of
the heliosphere. The effort, like the phenomena it seeks to describe, is
truly global in scale.

Mike Lockwood is a research scientist with the Space Science Department
of the Rutherford Appleton Laboratory and a visiting honourary lecturer
at Imperial College, London. Andrew Coates is a Royal Society university
research fellow, working at the Mullard Space Science Laboratory, in the
Department of Physics and Astronomy, University College, London.