Fifty years ago an Australian physicist called Bob Duncan reported that light from the Aurora Australis (or Southern Lights) was polarized. Although his discovery could have provided a new way of studying the atmosphere of Earth, other scientists at the time were unconvinced. Duncan's finding was quickly cast aside and the prevailing wisdom for the last half century has been that such light is not polarized. Now, however, an international team of physicists has made a similar measurement from the Arctic island of Svalbard, which suggests that Duncan was right all along.

The Aurora Borealis (Northern Lights) and Aurora Australis are spectacular displays of light that can be seen at high and middle latitudes. Aurorae occur when charged particles (mostly electrons and protons) ejected by the Sun are focussed and accelerated by the Earth’s magnetic field. These particles are believed to follow helical paths along Earth’s magnetic field lines.

Duncan considered what would happen when such twirling electrons collide with gas atoms about 200 km up in the atmosphere. He proposed that the electrons transfer energy to the atoms, leaving them in a specific excited quantum state. When they return to the ground state, the atoms emit light with a specific polarization through the process of fluorescence. And on one night in 1958, he measured a degree of polarization of 30% in light coming from the southern sky over Tasmania.

Jostled atoms

However, leading physicists of the day challenged Duncan’s finding because, in the process of emitting the light, the atom is expected to remain in the excited state for some time. During this time, it would be jostled about by other gas atoms, which — or so fellow researchers of the time argued — should destroy the polarization. As a result, Duncan’s observation was quickly discredited and forgotten.

Now, Roland Thissen and colleagues at the Planetary Science Laboratory at the CNRS Planetary Science Laboratory in Grenoble, France, and collaborators in the Netherlands and Norway have confirmed that auroral light is indeed polarized (Geophys Res Lett 35 L08804). The team focussed on red light emitted from fluorescing oxygen atoms and found that it has a maximum polarization of about 6% when it reached their instrument.

This figure was seen during in the “quiet” times between visible aurorae events when light is still being created even though it cannot be seen by the unaided eye. The polarization dropped to about 2% when an aurora was visible.

This drop, according to Thissen, means that Duncan’s critics were at least partially right. During visible aurorae the colliding electrons are thought to have higher energies than at quiet times. This means the electrons transfer more of their energy to the oxygen than at quiet times — and the process of giving up this extra energy before emitting red light tends to “smear” the polarization, said Thissen.

The observation was made using a standard optical technique involving splitting the light from the region of the aurorae into two channels, and passing one channel through a polarization filter. The polarization of the light can be determined by comparing the intensity of light in the filtered and unfiltered channels.

This is a standard technique that could be adapted to study the polarization of light from aurorae on planets such as Saturn and Jupiter. It could even be used to examine light coming from “exoplanets” orbiting other stars, says Thissen. This could help astronomers gain a better understanding of the both the magnetic fields and atmospheres surrounding these distant worlds.

Here on Earth, polarization studies could lead to a better understanding of quiet times in the aurorae and why the lights can suddenly flare up — often disrupting radio communications and other technologies.

Enigmatic measurement

Sadly, Duncan did not live to see his ideas revived — he died in 2004 — and Thissen said that it was unfortunate that they could not invite him to participate in their research. He said that Duncan’s measurement of 30% polarization remains an “enigma” because it represents the maximum polarization of light emitted from oxygen — something that can be seen in the lab, but should not be seen in light that is created in the atmosphere and then travels over 200 km before being detected.