The Earth’s magnetic field is, by every available measurement, the structural feature that protects the surface of the planet from most of the solar radiation that would otherwise sterilize it. The field is generated by the motion of molten iron in the planet’s outer core, approximately 2,900 kilometers below the surface. The field extends outward into space, where it deflects most of the charged particles streaming from the Sun, redirecting them around the planet rather than allowing them to reach the surface.
The standard cultural framing of the magnetic field treats it as a stable feature of the planet. The framing makes intuitive sense. The field has, on every available measurement, been present throughout the entire history of the planet’s habitability. The field has been, in some real way, one of the structural conditions that allowed life to colonize the planet in the first place.
What the standard framing does not, on the available evidence, fully absorb is that the field is not, in any meaningful sense, stable. The field has, more accurately, been reversing itself completely at irregular intervals throughout the planet’s geological history, with the magnetic north and magnetic south poles swapping positions. The reversals are well-documented in the geological record. The reversals have occurred, on the best current count, at least 183 times in the last 83 million years.
What the geological record actually shows
The evidence for the reversals comes from a particular property of certain rocks. When volcanic rock cools, the iron-bearing minerals within it align with the local direction of the Earth’s magnetic field, and remain locked in that orientation as the rock solidifies. The rock becomes, in some real way, a permanent record of the direction of the magnetic field at the moment of its formation. Sedimentary rocks, formed by the slow accumulation of magnetic particles in still water, preserve similar evidence.
By dating the rocks and measuring the orientation of their magnetic minerals, geophysicists have been able to construct, across the last several decades, a detailed timeline of the Earth’s magnetic field across hundreds of millions of years. The timeline documents that there have been at least 183 reversals over the last 83 million years, which works out to an average of approximately one reversal every 450,000 years. The intervals between reversals are not, on close examination, regular. Some periods of normal polarity have lasted only a few thousand years before reversing. Some periods have lasted tens of millions of years. The pattern, on the available statistical analysis, appears to be essentially random.
The most recent reversal is well-dated and well-studied. It is called the Brunhes-Matuyama reversal, after the two geophysicists who first documented evidence for it, and it occurred approximately 780,000 years ago. The reversal is sufficiently recent that the geological evidence is well-preserved and sufficiently old that no human ancestor of the modern species was around to witness it directly. The Homo sapiens species had not yet evolved. The hominin populations that did exist at the time were earlier species whose response to the event, if any, has left no documentary record.
What happens during a reversal
The structural picture of what occurs during a reversal has been refined considerably over the last two decades, and the picture is, on close examination, more dramatic than the standard cultural framing tends to credit.
The reversal does not, in most cases, happen quickly. The most recent careful analyses, based on global data compilations of paleomagnetic records, suggest that the Brunhes-Matuyama reversal took approximately 22,000 to 30,000 years to complete from start to finish. Research published in the Journal of Geophysical Research in 2023 placed the duration at approximately 30,000 years, with the main polarity transition occurring within a narrower window of approximately 10,000 years centered on 778,000 years ago.
What happens during this period is, on the available evidence, structurally significant. The strength of the magnetic field, in the early phase of the reversal, decreases substantially. The decrease is not partial. The field can weaken to as little as 10 percent of its normal strength. The weakening is the structural mechanism by which the reversal occurs. The dipole component of the field essentially collapses, the field becomes briefly disorganized, and then the dipole reorganizes itself in the opposite orientation.
The implications of the weakening are considerable. The magnetic field is, in normal operation, the primary defense against solar radiation reaching the planet’s surface. When the field is at 10 percent of normal strength, the defense is correspondingly reduced. The planet is, during the transition period, more exposed to solar radiation than it is during periods of normal polarity. The exposure is not, on the available evidence, sufficient to sterilize the surface. The exposure is, however, sufficient to produce measurable changes in the atmosphere, in the production of cosmogenic isotopes such as carbon-14 and beryllium-10, and in the geographical distribution of various atmospheric phenomena.
What the auroras look like during a reversal
One of the more striking features of magnetic field weakening is what happens to the auroras. Under normal conditions, the auroras occur in roughly circular regions around the magnetic poles, called the auroral ovals. The ovals are produced by the interaction of solar wind particles with the magnetic field, and they are positioned at high latitudes because the dipole field channels the particles toward the polar regions.
When the dipole field weakens, the auroral ovals expand. The expansion occurs because the field is no longer concentrating the solar particles as effectively at the poles. The particles can interact with the atmosphere at lower latitudes than they normally would, producing auroras in regions where, under normal conditions, they would essentially never appear.
This is, on close examination, what would have been visible during the Brunhes-Matuyama reversal, had any conscious observer been present to witness it. The auroras would have been appearing not only at the high latitudes where they normally appear, but at middle and possibly even tropical latitudes. The night sky over the equator, during the weakest periods of the transition, may have been illuminated by aurora-like phenomena that the modern equivalent of the equator simply does not produce.
The evidence for this is indirect but consistent. Geological and atmospheric records from the period show signatures of increased atmospheric ionization at lower latitudes, which is structurally consistent with the kind of auroral activity that the weakened field would have produced. The exact details of what the sky would have looked like to a hypothetical observer are not recoverable from the geological record. The general structural picture, however, is clear.
What this implies for the present
The Earth’s magnetic field has, on the available measurements, been weakening over the last several centuries. The field is currently about 9 percent weaker globally than it was 200 years ago. The weakening is most pronounced in a region called the South Atlantic Anomaly, where the field strength is considerably lower than the global average and has been declining at an accelerated rate.
The wider research community has not, on the available evidence, reached a settled view on whether the current weakening is the early stage of a developing reversal. The arguments on each side are worth attending to. The weakening is real and ongoing. The pattern of geomagnetic excursions in the past has sometimes been the precursor to a full reversal and has sometimes resolved without one. The current observations are not, on the available evidence, sufficient to distinguish between these possibilities.
What is clear, on close examination, is that the next reversal will, at some point, occur. The pattern of 183 reversals across 83 million years suggests that reversals are, in some real way, the normal long-term behavior of the field, with periods of stable polarity being the intervals between them. The current period of normal polarity, which has lasted 780,000 years, is somewhat longer than the average interval but not dramatically so. Whether the next reversal occurs in the next thousand years, the next ten thousand years, or the next hundred thousand years is not, on the available evidence, currently predictable with any precision.
The acknowledgment this article wants to leave
The Earth’s magnetic field is, by every available measurement, considerably less stable across geological timescales than the standard cultural framing has been treating it as. The field has reversed itself completely at least 183 times in the last 83 million years. The most recent reversal occurred approximately 780,000 years ago and took somewhere between 22,000 and 30,000 years to complete. During the transition, the field’s strength dropped to a small fraction of its normal value, the planet was more exposed to solar radiation than under normal conditions, and auroras likely appeared at latitudes where they essentially never appear today.
The pattern is well-documented. The pattern is, on the available evidence, ongoing. The current period of normal polarity is, by structural design, one of the intervals between reversals rather than the permanent condition of the field. The next reversal, whenever it occurs, will involve the same general structural features. The strength of the field will decrease. The planet will be briefly more exposed to solar radiation. The auroras will spread to latitudes where the modern population has never seen them. The pattern, on the available geological record, is what the field does. The current configuration is, more accurately, the temporary state between events that the planet has been performing throughout its history, and that the wider cultural register has not yet fully absorbed as the structural feature it is.