>

© H. Raguet/CNRS Photothèque
To understand the magnetic field, the DTS experiment uses a small-scale model of the Earth's core: two nested spheres (inner and outer core) separated by hot liquid sodium, that are then rotated rapidly around a vertical axis.
Reversals
We now know that the poles have reversed more than 300 times in the last 200 million years. Prior to each reversal, the magnetic field weakens, causing the magnetic shield to become less intense. “The cycles are extremely irregular,” Hulot explains. “Periods of stability lasting 30 million years have been observed, as have reversals at intervals of only a few 10,000 years. If there’s a pattern, it must be probabilistic and very subtle.” The most recent reversal took place 780,000 years ago. We are the living proof that our ancestors–Homo erectus–survived it. On the other hand, if a reversal occurred today, satellites and electrical power grids would suffer frequent disruptions since the Earth would no longer be sheltered from magnetic storms–these sudden surges of charged particles caused by eruptions on the Sun’s surface.
To better understand the Earth’s magnetic field, scientists are studying other planets, like Mars, which has kept traces of a magnetic field that has long since vanished. But they are also attempting to recreate this phenomenon in the laboratory. For instance, geophysicists have managed to reproduce a dynamo mechanism–though still considerably different from the one operating at the center of the Earth. “We use liquid sodium, which provides the best electrical conductivity and the best fluidity,” Nataf explains. “To reproduce the Earth’s magnetic field,” Hulot adds, “we would need a much more conductive and fluid material. We just don’t have one. But we learn a great deal from these experiments.”
In the last few hundred years, the intensity of the field has fallen by 10%. Is another reversal coming? “It could simply be a minor disturbance,” says Hulot. “No deduction can be made about when the next reversal will take place.” We can only estimate variations in the magnetic field over the next five years–already quite an achievement.
Denis Delbecq
1. Institut de Physique du Globe de Paris (CNRS / IPG Paris / Universités Paris-VI and VII / Université de la Réunion).
2. Laboratoire de géophysique interne et tectonophysique (CNRS / Universités Grenoble-I et de Chambery / Laboratoire central des Ponts et chaussées / IRD).
3. Observatoire des sciences de l'Univers de Grenoble (CNRS / Université Grenoble-I / Institut national polytechnique de Grenoble / IRD).
Gauthier Hulot,
gh@ipgp.jussieu.fr
Henri-Claude Nataf,
henri-claude.nataf@obs.ujf-grenoble.fr