The idea that a record of the Earth’s magnetic past might be stored in objects made from fired clay dates back to the 16th century. William Gilbert, physician to Queen Elizabeth I, hypothesized in his work De Magnete that the Earth is a giant bar magnet and that clay bricks possess a magnetic memory. This phenomenon – known as “thermoremanent magnetization” – now forms the basis of a well-established method for dating archaeological sites that contain kilns, hearths, ovens or furnaces.

Indeed, the study of these burnt materials containing magnetic minerals, found at archaeological sites, is known as “archaeomagnetism”. One of the aims of this field is to help geophysicists gain a better understanding of local changes that have occurred in the Earth’s magnetic field over the past 3000 years. And if we know how the field changed in the past, we could also get insights into our magnetic future.

We already know that the Earth‘s magnetic field has lost around 10% of its intensity over the last 150 years. “The dipole strength has been steadily decreasing at a rate such that, should it continue, in 2000 years the magnetic field strength would be zero,” says geologist Rory Cottrell of the University of Rochester in the US. “The thought is that the planet is headed for a magnetic field reversal.”

Data collected from magnetic records in rocks indicate that over the last 76 million years there have been 170 reversals in which the north–south polarity of the field has completely switched. And it seems that another event is overdue: a full reversal has happened, on average, every 200,000 years over the last 10 million years, and the last one was 780,000 years ago.

Even small changes in the Earth’s magnetic field can have far-reaching repercussions for the planet’s surface. That’s because the magnetic field acts as a shield, repelling and trapping charged particles from the Sun that would otherwise cause electrical grid failures, navigational system malfunctions, and satellite breakdowns. Strong solar winds already cause problems from time to time, notably in 1989 when a billion-tonne cloud of solar plasma breached the Earth’s magnetic field. This created electrical currents in the ground that caused an electrical power blackout across the entire province of Quebec, Canada. If the field weakens further, we can expect more such events, triggering major disruptions.

A particular cause for concern is the South Atlantic Anomaly, an area stretching from Chile in South America to Zimbabwe in Africa, where the magnetic field intensity is much lower than the global average (figure 1). The magnetic field strength across this region can reach as low as 25 μT, compared with up to 67 μΤ for other parts of the Earth’s surface. “It’s low enough that incoming radiation is no longer deflected and it interferes with satellite transmissions,” says Vincent Hare, a geophysicist and archaeological scientist at the University of Cape Town, South Africa. What’s more, the anomaly has been growing and intensifying over the last 200 years or so, which could be yet another signal that a field reversal is on its way.

Measuring archaeomagnetism

To understand whether this localized anomaly might be a sign of more significant changes, geophysicists have been examining our planet’s relatively recent magnetic history. Unfortunately, direct observations of the Earth’s magnetic field have only been collected since the 1850s, and even then only in some locations. While magnetic information contained in rocks goes back millions of years, researchers have focused their attention on archaeological artefacts to reconstruct our magnetic history over the last 3000 years or so.

Clays or other materials containing magnetic minerals lose any magnetic ordering when they are heated above 570 °C, but then become imprinted with the Earth’s ambient magnetic field when they are cooled

The field of archaeomagnetism relies on the fact that clays or other materials containing magnetic minerals – usually magnetite – lose any magnetic ordering when they are heated to above 570 °C (the Curie temperature). Indeed, the sample loses its net magnetization but when it cools back down below the Curie temperature, the particles remagnetize in the direction of the local magnetic field at that time. In this way, these archeological samples provide a snapshot of the Earth’s ambient magnetic field through different times and places in history. These samples can reveal both the intensity of the magnetic field and its direction, which is measured at any point on the Earth’s surface by the declination – the angle on the horizontal plane between magnetic north and true north (figure 2).

The first attempt to extract magnetic information from fired clay was made in the late 19th century, when the Italian scientist Giuseppe Folgheraiter calibrated a “geomagnetic secular variation curve” – a record of changes in both the declination and inclination of the Earth’s magnetic field, in a given location – for dating ancient pottery. The technique became more established in the 1970s, and can now deliver the same sort of precision as radiocarbon dating. “The job of the archaeomagnetist is to take samples and measure them to death,” says Hare.

But obtaining accurate measurements is far from simple. For a start, the residual magnetism in archaeological samples is tiny, with magnetic moments in the order of 10^–3–10^–5 Am²/kg, which is an order of magnitude lower than would be required to move a compass needle. Such small magnetic signals can only be detected with cryogenic magnetometers made from superconducting quantum interference devices (SQUIDs). Experiments must also be carried out in a “magnetic vacuum”, often using a Helmholtz cage that creates a uniform magnetic field to cancel out the Earth’s magnetic field.

Another complication is the compound nature of the raw magnetic signal. “The measurements are often a vector sum of the ancient magnetization you’re interested in, and also more recent overprints,” says Andy Biggin, a palaeomagnetist at the University of Liverpool in the UK. Those more recent or “secondary” magnetizations, he says, can often be successfully removed by incrementally heating the samples to temperatures approaching the Curie point, and then cooling them down after each heating step. “That gradually strips away the less stable magnetizations,” Biggin adds.

The accuracy of the measurements also depends on the magnetic structure of the sample, with smaller magnetic grains retaining their magnetization for longer. In magnetite, for example, magnetic grains measuring 50–80 nm can store their magnetic information for billions of years. Obviously, accurate readings of the field direction can only be taken if the object has not been disturbed since the field was imprinted. In other words, the materials should not, for example, have been significantly dried or water-logged since the sample was last heated to the Curie temperature. This requirement rules out a number of types of samples including those that may have been disturbed while being found. “Pottery kilns, or even [the remains of] cities that burned down, make an excellent archaeomagnetic record,” says Biggin.

It is even harder to determine the magnetic intensity from archaeological artefacts, since the present-day measurement also depends on the intrinsic ability of the sample to acquire thermoremanent magnetization. The easiest way to determine this intrinsic property, says Biggin, is to expose the sample to a known magnetic field and then measure the resulting magnetization. If, for example, the new magnetization is twice as strong as the ancient magnetization, the ancient magnetic field must have been half as strong as the controlled field used in the lab.

But working with ancient pieces of clay inevitably introduces problems, partly because the heating process often causes chemical changes or physical deterioration in the samples. “I have never encountered an analytical technique that is so difficult and takes so long,” says Hare. “You can get a month into your measurements, and then have them fail.” Despite these difficulties, geophysicists have already pieced together an accurate magnetic record for western Europe and large parts of the Middle East. In the UK, for example, archaeomagnetic dating now extends back to 1000 BCE, in some cases with accuracies within tens of years.

 

Searching for anomalies

This increasing amount of archaeomagnetic data suggests that the current South Atlantic Anomaly is not the only example of extreme local variations in the recent history of the Earth’s magnetic field. One area of focus is the Middle East, where a team of geologists and archaeologists has been studying the magnetization of ancient artefacts found in Israel. “We found very mysterious magnetic fields and surprisingly different than expected – super super strong,” says geophysicist Ron Shaar from the Hebrew University of Jerusalem. Strange behaviour was seen in the field direction as well as the intensity, with anomalies of more than 10° from the prevailing field direction of the time.

In 2016 the team published results obtained from pottery shards and cooking ovens found at Tel Megiddo and Tel Hazor, two sites in Israel that were occupied during the Iron Age more than 3000 years ago (figure 3). The data reveal the evolution of an extreme geomagnetic high between the 11th and 8th centuries BCE, culminating in two “archaeomagnetic jerks” or “geomagnetic spikes” where the field intensity shoots up and down again in less than a century (Earth and Planetary Science Letters 442 173). The two spikes are centred at 732 BCE and 980 BCE, and each one has a field strength more than twice that of the current dipole field. This period has since become known as the Levantine Iron Age Anomaly.

One intriguing possibility is that the effects of these high fields might have been seen during biblical times. Passages from the book of Ezekiel, written 2600 years ago and chronicling a journey through Turkey, describe an immense cloud with flashing lightning surrounded by brilliant light. This depiction is thought to refer to the Aurora Borealis, normally only observed in the far north when charged particles collide at high speed with the stronger magnetic field in these regions. But a stronger magnetic field over the Middle East at that time could explain the lights seen by the prophet. Although the time period doesn’t match exactly with the spikes detected by the team, geophysicist Amotz Agnon – who  founded the paleomagnetism lab in Jerusalem and initiated the Tel Megiddo project – points out that “with prophecies, you never know, maybe [this was] just a rumour from some oral tradition”.

In 2020 Shaar and his team published new data from an Iron Age excavation site in Jerusalem. They analysed 397 samples of burnt material from the floor of a building that they deduced was destroyed during the Babylonian conquest of the city, dated to August 586 BCE. Their results reveal similar high-field values, and also provide an exact anchor date for their measurements (PLOS ONE 15 e0237029).

Data from further afield show just how far the anomaly stretched at that time. “We can trace its evolution, how it starts in the Middle East and migrates westward toward western Europe over a period of a few hundred years or so,” says Shaar. He and others have studied material from Turkey and Cyprus that show large swings in magnetic field direction from 1910 to 1850 BCE, with exceptionally high intensities around 700 BCE. Other data from Georgia showed high-field values in periods stretching from the 10th to 9th centuries BCE, as well as fast-field variations about 500 years later.

But this unusual magnetic behaviour is quite different from the anomaly now seen in the southern Atlantic, which is a localized region of weak magnetic field. To find examples of similar low-intensity anomalies from the past, Cottrell decided to search for clues in southern Africa. Working in collaboration with South African archaeologists, Cottrell identified suitable samples from sites near the Shashe and Limpopo rivers in northern South Africa, Mozambique, Botswana and Zimbabwe, dated from 425 to 1550 CE.

“The Iron Age of southern Africa is a good place to go,” says Hare, who was part of the study team. He explains that the local people would have built huts with clay floors, and regularly performed certain rituals to cleanse the community if there was a drought or a similar event. “One of those would be the burning of a hut floor, and that’s perfect for archaeomagnetism.”

So far only field-direction measurements have been published, but these early results already show interesting anomalous behaviour (Geophysical Research Letters 45 1361). “If we look at the magnetic field between today and 1500 CE, the rate of change was on the order of 0.06° per year, but between 1500 CE and 1350 CE, it was almost double that,” explains Cottrell. The team also identified an earlier period of relatively rapid change between the 6th and 7th centuries CE.

Cottrell believes that this variability is the most recent historic display of whatever phenomenon is causing the current South Atlantic Anomaly. This had previously been thought to be only a very recent event, but these new findings suggest that some parts of the world might to be prone to repeated changes in the magnetic field. To test this idea, Biggin looked further back in the geological record. He studied volcanic glasses formed 8–11.5 million years ago on the island of Saint Helena, right in the middle of the South Atlantic Ocean, and also found large variations in the direction of the magnetic field. This finding therefore supported the view that the Earth’s magnetic field has been unstable in this region for millions of years.

Under the mantle

It is still unclear why certain regions experience these continuing anomalies, but geophysicists believe that the answer may lie in the interactions between the Earth’s mantle and its outer core, the 2889 km layer of molten iron-rich rock that is responsible for the Earth’s magnetic field. The magnetic field is generated by a dynamo process in which the Earth’s rotation, combined with convection currents in the molten core, creates rotating columns of liquid that generate the magnetic field. “When you move a conductor through a magnetic field, you induce electric currents and that makes more magnetic field – so it’s self-sustaining,” explains Biggin.

Anomalies in the magnetic field are thought to be associated with patches of magnetic field in the outer core that are stronger or weaker relative to the overall magnetic dipole, or that even point in the opposite direction. “As these flux patches move, they intensify and diminish, and cause very fast local changes,” says Biggin. Indeed, the current South Atlantic Anomaly seems to sit on top of one or more patches of opposing flux.

The Rochester team has proposed that these flux patches are associated with temperature or density changes deep in the Earth’s mantle. “Africa sits on top of a very special seismological feature in the interior of the Earth, called a large low-shear-velocity province,” says Hare. “It’s essentially a slightly heavier portion of the lowermost mantle of the Earth that sits on top of the outer core, and protrudes slightly into it.” This protrusion then perturbs the flow of the liquid outer core, causing flux patches that alter the magnetic field on the Earth’s surface.

Future clues

Overall, the data from archaeomagnetic studies have been reassuring for the future of the Earth’s magnetic field, since the anomalies we see today are clearly in line with past behaviour. “What we have observed over the past several hundred years is a very normal behaviour of the geomagnetic field,” says Shaar. “There is nothing to worry about based on comparison of today’s field with what we know about the ancient field.” That view is backed up by magnetic records obtained from rocks that were formed much further back in the Earth’s history, which show that the Earth’s magnetic field is now globally much stronger than in the 50 000 years leading up to the past five reversals.

Despite this general reassurance, there is still a lot to explore and understand about the anomalies in the Earth’s magnetic field. That means collecting more data on intensity variations over the last three millennia, but that’s a daunting task when there is still such a high failure rate in sample analysis. “Very few [geophysicists] focus on intensity, because the experiments drive them mad,” says Hare. “But it’s the key to this whole question.”

A new measurement technique being developed combines computed X-ray tomography with scanning magnetometry

One solution could be a new measurement technique being developed by Lennart de Groot, a geophysicist from Utrecht University in the Netherlands. Rather than simultaneously measuring the magnetism of millions of grains in one sample, de Groot combines computed X-ray tomography with scanning magnetometry to calculate the unique contribution of each grain (Geophysical Research Letters 45 2995). More accurate results can be achieved, he says, because the technique requires only a small subset of the magnetic grains contained within each sample.

It will also be important to source samples from a wider variety of locations. Detailed magnetic profiles now exist for Europe and much of the Middle East, while data coverage in China is also improving. It remains difficult to source magnetic data from the southern hemisphere, but Cottrell is continuing her work in Africa, adding that “there has been a concerted effort by many researchers, particularly from South America, to collect this data”.

As more data become available, geophysicists are convinced that more anomalous behaviour will emerge in the Earth’s magnetic history. There is already some evidence of strong flux patches under Siberia, for example, and in the Southern Ocean near Australia. Shaar agrees that other anomalies will be found, and predicts that any new discoveries will be just as puzzling as those reported so far. “The world is huge and I suspect that we will find in the future that the geomagnetic field is nothing like we have measured in the past few hundreds of years,” he says. “It is an evolving thing, constantly changing, and there will be many surprises.”