Mars no longer has a significant global magnetic field. But we now know that it had one early in its history. Observations by NASA's Mars Global Surveyor (MGS) spacecraft have revealed the tell-tale signs in the ancient crust of the southern hemisphere: stripes of magnetised crust of alternating polarity running roughly parallel for 2000 km.
The strength and structure of the magnetism has taken planetary geologists by surprise. Similar stripes discovered on the Earth's seafloor in the 1960s were pivotal in gaining acceptance for the theory of plate tectonics. Do the new findings mean that plate tectonics also operated on early Mars? Until now, the MGS data geologists thought it hadn't, but now they're not so sure.
A benefit from misfortune
The discovery was made thanks to a couple of unfortunate incidents which resulted in MGS going closer to the Martian surface than originally intended. The size and significance of the magnetic stripes probably wouldn't have been appreciated from a greater observing height.
"After the loss of the Mars Observer spacecraft in 1993, NASA decided to build a new mission, MGS, that was cheaper, faster and better. To make the spacecraft cheaper, we needed to save weight and fuel," explains Henri Rème from the Centre d'Etudes Spatiales des Rayonnements, Toulouse, France, who is an investigator on the MGS magnetometer instrument. Mars Observer had a powerful on-board engine to inject it into the correct circular orbit. With MGS, it was decided to save weight by using a smaller engine and the friction of the atmosphere to nudge the spacecraft gradually into its final orbit.
During this aerobraking phase, part of the MGS orbit was just over 100 km above the Martian surface, closer than the 350 km height of MGS' final orbit or that planned for Mars Observer. For a few orbits, the close passes revealed disjointed areas of magnetism in the crust. "Then the second bad thing happened that was good for us," says Reme. "The arm of one of the solar panels wasn't standing up well to the friction, so NASA decided to take the aerobraking process more slowly. Aerobraking would normally take three months, but it was decided to have a long phase lasting 15 months. That meant there were a lot of low altitude passes."
A pattern emerges
Gradually a pattern began to emerge from the magnetometer data. 100 km wide stripes of magnetised crust running roughly parallel for up to 2000 km, were revealed in the southern hemisphere. The stripes alternate in polarity with the north magnetic pole of one pointing up from the surface and the north magnetic pole of the next pointing down.
"About 12 of us were sitting round a table at the Jet Propulsion Laboratory in Pasadena, California before the first low altitude pass," says Rème. "We decided to place a bet on the size of the magnetic field. Our predictions ranged from 5-100 nT. At the first pass, MGS recorded a value of 55 nT. Two days later it recorded 400 nT, which was a complete surprise. But when the spacecraft went above the southern hemisphere, the surprise was fantastic. We obtained 1600 nT."
When molten rock containing magnetic material, such as haematite (Fe2O3), cools and solidifies in the presence of a magnetic field, it becomes magnetised and takes on the polarity of the background field. This magnetism is lost only if the rock is subsequently heated above a particular temperature (the Curie point which is 770oC for iron). The magnetism left in rocks is therefore a record of the magnetic field prevailing when the rock solidified.
An ancient magnetic field
Most of the rock in the southern hemisphere of Mars solidified about 4 billion years ago. Younger rock, or old rock that has been subsequently re-heated (by impacts for example), shows no magnetism. This indicates that the global field lasted for only the first few hundred million years of Mars' life, when the temperature of the molten iron in the planet's core might have been high enough to churn it into a magnetic dynamo. "Why did the Martian magnetic field disappear? We need to know," asks Rème.
The stripes suggest that the field changed polarity in a similar way to the Earth's magnetic field - several times per million years, but not regularly. When similar stripes were discovered on Earth in the 1960s, they were taken as evidence of plate tectonics, the process by which vast slabs of the Earth's crust gradually move over underlying magma. The stripes were found in the ocean floor, spreading out from either side of underwater vents in the crust where magma wells up continuously to form new seafloor that gradually pushes the continents apart.
There are some differences, however, between the magnetic stripes on Earth and those on Mars. The Martian stripes are wider, much more strongly magnetised and do not appear to spread out from a central crustal spreading zone. "There is not enough data to know for sure whether there was plate tectonics or not," says Rème, who is now involved in planning a micromission to collect more data. "Geologists think plate tectonics never existed on Mars. But we now know it's a possibility."