It is not only rocks that geologists are using to track the earth's constantly changing magnetic field. Geoff Watts reports on how centuries-old measurements are playing a part.

In the early 18th century, the King George was one of a fleet of British East India Company merchant vessels making the arduous but routine voyage to Asia. Sweating it out in the midsummer heat of the Indian Ocean, the King George's captain might have been surprised to learn that the daily compass measurements he was recording in the ship's log would be scrutinised 300 years later by researchers at Leeds University.

On a geological time scale, the nautical compass data that earth scientist Andy Jackson and historian Art Jonkers have been collecting is almost absurdly small. Their colleagues, such as Dave Gubbins, also of the Leeds school of earth sciences, are more likely to use measurements derived from rocks laid down over many millions of years. What all three researchers have in common is an interest in the earth's magnetic field.

Rocks take on the direction of the earth's field at the time they are laid down. And they stay that way, offering geologists a permanent record of the changes that have taken place over millions of years. Although the field is constantly changing - with the relative positions of geographical and magnetic North depending on time and place - scientists can trace its history. Professor Gubbins has recently proposed that its changes of direction are not random, but are the product of a complex relationship between the solid and liquid regions of the earth's core.

ÁñÁ«ÊÓƵ

The changes are also more extensive than most people realise. "In the past, the field has often changed completely, with magnetic North becoming South," says Professor Gubbins. "The strength of the field falls away for 5,000 to 10,000 years. Then comes the actual change of direction, which takes perhaps 1,000 years, and finally the field slowly builds up again."

The magnetic field is generated by the movement of molten iron in the liquid portion of the earth's core. "As you descend through the earth, you meet the liquid core about halfway down," says Professor Gubbins. "It moves at a few millimetres per second - fast by geological standards. In fact, liquid iron can pass from the bottom of the core to the top in a few hundred years." At the very centre, over a diameter of about 2,400km, the pressure is high enough to solidify the iron.

ÁñÁ«ÊÓƵ

Professor Gubbins is aiming to explain why, even when the poles have wandered as far as the Equator, they usually return without changing places. His argument, in essence, is that the magnetic field of the liquid region can change more rapidly than that of the solid, the latter exerting a kind of damping effect.

"It's quite likely that without a solid core, the earth's magnetic field would reverse more frequently," he says. "The inner core has what's termed magnetic inertia."

Every so often one of the polar wanderings goes so far and is so sustained that the solid core has time to catch up. "The present position, with the North magnetic pole in the geographical North, has lasted 750,000 years. In that time we've had ten or 20 of these excursions in which the pole has moved a long way and then gone back. That fits rather well with the idea that the liquid core can change in 500 years while the solid part of the core takes 5,000 years - ten times longer. You can expect ten of these excursions to every complete reversal."

Since Roman times the magnetic field has fallen by 40 per cent. A further thousand years would bring about another polar reversal. And, says Professor Gubbins, we are indeed due for one. "The last was 750,000 years ago, and we've usually had two or three per million years. On the other hand, the magnetic field has dropped like this many times during the past 10,000 years and gone back up again."

ÁñÁ«ÊÓƵ

Predicting magnetic field change is difficult. Even the largest supercomputers are not up to it. But when another reversal does happen, it will have all sorts of effects, from atmospheric disturbances to difficulties in radio communication. We would also lose the partial shield against cosmic rays provided by the earth's field.

If Professor Gubbins can be thought of as painting magnetic murals, Andy Jackson and Art Jonkers are at the miniature portrait end of the business. They use measurements found in the log books of ships dating from the early 1600s, mostly sailing between European countries and their colonies.

"Sailors typically measured declination, the angle between true North, which they got from the sun, and magnetic North. They would measure that every day or sometimes twice a day, and note it down in the log. They would have to do that to navigate accurately. Many of these logs have been preserved in archives around Europe, especially the British Library."

Doctor Jackson and Mr Jonkers have been impressed by the accuracy of the records. "They could measure declination to better than half a degree. We've now got a tremendous data set - almost 170,000 readings." Mr Jonkers, who also works as a researcher at the Free University of Amsterdam, is interested in the history for its own sake. But Doctor Jackson wants to create a model of the earth's field over the past 400 years: more data by which to understand the behaviour of what even Albert Einstein rated as one of the greatest mysteries.

ÁñÁ«ÊÓƵ