The people, science and technology behind discovery

Subscribe to the Magazine

Get new articles sent directly to
your mailbox.

Join the network

Twitter Updates Group Forum

Earth Explorer is an online source of news, expertise and applied knowledge for resource explorers and earth scientists. Sponsored by Geosoft.

Upcoming Events

SERDP and ESTCP Symposium

November 27 - 29, 2018

AGU 2018

December 10 - 14, 2018

News & Views

All News

Rapid Reversal in the Earth’s Magnetic Field

Geoscientists uncover an example of rapid transitional field change (RTFC)
in Nevada’s Sheep Creek Range

Figure 1. Equal-area plot showing site-mean remanence directions (and 95% confidence limits) from flows near the top of the Sheep Creek transition zone. Open (closed) symbols correspond to upward (downward) pointing directions. Flow 20 is bracketed by Flow 19 (below, with E-down paleomagnetic direction) and Flow 21 (above, with Ndown direction).

[Click to enlarge]

Figure 2. Stepwise demagnetization of samples from a vertical profile through Flow 20. (a)Vector endpoint diagrams (horizontal component only) for thermal demagnetization on samples 1.2 m, 1.7m, 2.7m, and 2.9 m above flow base. Labels identify remanence components in the E-down direction of Flow 19 and N-down direction of Flow 21. Above 175°, heating steps range from 25°C to 10°C. Highest temperature steps are all 585°C. Plots are scaled so that north components of NRMs are equal. (b) Vector endpoint diagrams like those in (a) for AF and thermal demagnetization of companion specimens from 0.6 m above flow base. (c) Tub distributions for the N-down (black) and E-down (light gray) components of samples from 1.2m ,1.7m, 2.4m, and 2.7m above base of Flow 20 (overlapping Tub ranges shown in dark grey.) Height of bars is proportional to rate (per °C) of moment loss, with maximum rate at each level normalized to 1. Bars extend from lower to upper Tub of remanence component unblocked. Arrows indicate picks for baking temperature derived from these plots.

Figure 3. Initial susceptibility versus temperature (in argon) for sample 1.4 m above base of Flow 20. Sample was cycled to target temperatures of 150°C, 250°C, 300°C and 350°C before final heating to 625C. Arrows identify heating and cooling segments of curves. Curve for heating cycle to 250°C obscured by lower temperature portion of the 350°C curve. No progressive increase of Curie temperature during the heating cycles (diagnostic of low-temperature oxidized titanomagnetite) is displayed by the sample.

[Click to enlarge]

Figure 4. Results from conductive cooling model. (a) Time evolution of temperature profile in 3.9 m thick flow cooled to 150°C then capped by 8.2 m thick flow at 1100°C. (b) Inferred maximum baking temperatures in 3.9 m thick flow baked by 8.2 m flow. Curve with dash-dot pattern (1) shows temperature profile immediately before emplacement of upper flow. Solid curves show maximum baking temperatures if lower flow is fully cooled (2) or conductively cooled to maximum temperature of (3) 125°C, (4) 150°C, and (5) 200°C before upper flow is emplaced. Squares show maximum baking temperatures (corrected for cooling rate) inferred from the paleomagnetic data from Flow 20. Error bars show range of sample estimates. (c) Time evolution of temperature 1 m above the base of 3.9 m flow baked by 8.2 m flow. Remanence acquisition in E-down direction occurs between times I and II and all or part of the interval between II and III. Between III and IV, remanence previously acquired between II and III is unblocked. Remanence acquisition resumes (in N-down direction) starting at IV. Change in field direction occurred between II and IV. (d) Overall constraint on duration of rapid field change. Shaded area corresponds to the interval between times II and IV on (c).

By Leonard Chan

Since the 16th century, we have known that the Earth is a giant magnet with a dipolar magnetic field. In the 1920s, we learned that the magnetic field undergoes reversals, exchanging its North and South poles at numerous times throughout our planet's history. The quest to understand the cause of these geomagnetic reversals is ongoing, as geoscientists scour the Earth for evidence of past field reversals, recorded in ferromagnetic minerals of solidified sedimentary deposits or volcanic flows cooled on land.

Towering 9,733 feet over the barren playa of the Alvord desert, the east face of the Steens Mountain is a journey back through time. Composed of basalts stacked one atop the other, each lava flow offers a snapshot of the earth's magnetic field. In that stack is recorded a 16.7 million year old polarity transition. When, in 1995, geoscientists examined one lava flow in particular, a 4 meter thick flow erupted during the transition, they found that the top and base, the parts that cool quickest, were magnetized in directions different from its interior, a pattern possible only if the geomagnetic field changed 6°/day. This is astonishingly high and orders of magnitude greater than typically-observed rates of secular variation.

The timeline for geomagnetic field reversals is on the order of a few thousand to a few tens of thousands of years with the magnetic direction believed to reorient slowly over this time frame.

These previously-held beliefs are now being challenged by a new hypothesis called "rapid transitional field change" (RTFC). Developed by Scott Bogue of Occidental College and Jonathan Glen of the United States Geological Survey (USGS), this controversial hypothesis suggests that transitions can be punctuated by rapid directional field changes many orders of magnitude greater than the steady changes observed in historical times or those typically observed during reversals.

The RTFC hypothesis has been debated, with some geoscientists arguing that the Earth's liquid core is incapable of generating the magnetic fields required to induce such rapid change, and others who contend that the electrical conductivity of the lower mantle would block such high-frequency electromagnetic signals from being observable at the Earth's surface.

And perhaps more importantly, up until now, the only evidence to support the hypothesis was from Steens Mountain.

But Bogue and Glen believe they have uncovered a second example of this phenomenon.

Nestled in the Sheep Creek Range in Lander County, Nevada, is a wellspring of paleomagnetic data: an exposed 150-meter thick section of basaltic and basaltic andesite lava flows. The radiometric age of these lava flows is 15.58 million years, approximately 1 million years younger than the flows at Steens Mountain.

One flow in particular caught the scientists' attention. Dubbed Flow 20, this 3.9m thick lava flow was sandwiched between two other flows, appropriately named Flow 19 (below) and Flow 21 (above). Using a portable gas-powered drill (essentially a converted chainsaw) armed with a diamond-studded bit, 1"-diameter cylindrical cores were extracted from the flows and cut into 1”-long specimens for analysis.

All the lava flows in the Sheep Creek Range acquired their magnetization, otherwise known as remanence, during the later stages of a reverse-to-normal polarity transition. Each of the flows had a primary remanence, acquired when the rock was formed and generally indicative of the direction and strength of the earth's geomagnetic field at the time, and a secondary remanence, which can be the result of more localized effects, such as lightning strikes or influences from the present-day field. In order to isolate the useful primary remanence, the secondary remanence was stripped away using techniques such as stepwise alternating-field (AF) or thermal demagnetization.

"Thermal was nice in this case because it essentially mimics, but in reverse, the acquisition process. When we heat the sample up, we do so in a stepwise fashion to incrementally demagnetize the sample. The remanence lost in a given temperature range is equivalent to what was gained at those temperatures. Those grains that are the least stable, whose magnetization is easily lost, do so at low temperatures. The more stable magnetizations are only removed at the higher temperatures. By doing this, we can watch how it changes, and determine precisely the different components," explains Glen.

After each demagnetization cycle, the samples were analyzed with a cryogenic magnetometer, revealing an unusual acquisition history.

Bogue and Glen hypothesize that Flow 20's initial direction was acquired when it first erupted over top of Flow 19. However, Flow 21 erupted shortly afterwards, reheating the underlying Flow 20. The uppermost part of Flow 20 was reheated to such a degree that it effectively erased the memory of its original remanence, and acquired a different magnetic direction as it cooled for a second time.

The difference between the two magnetic directions recorded by Flow 20 was estimated to be 53°, indicating a significant change in the Earth's geomagnetic field from Flow 20's initial eruption to its final cooling.

While they knew the change was substantial, what they did not know was the period of time over which it occurred. Using an open source programming language, they coded a one-dimensional model of the cooling history of a 3.9m thick lava flow that was capped by an 8.2m thick lava flow at 1,100°C.

Their results indicated that the remagnetization could only be accounted for if Flow 21 erupted while Flow 20 was still cooling. They estimated that the observed geomagnetic changes happened over the period of a year, translating to a little over 1°/week. While not as high as those recorded at Steens Mountain, this rate of field change is still 2-3 orders of magnitude greater than that of typical secular variation.

Glen says that there is still much to learn about reversals and admits the possibility that the RTFC phenomenon may be localized. "The reversal process is more complex than these rapid directional changes. This is one place in the world where we've observed this. The other is Steens Mountain. We don't know how extensive these kinds of fluctuations are around the globe. There may be localized eddies. Smaller currents produce fields only locally, which you might not see on the other side of the globe. During a transition, these local features might dominate."

Regardless of the mechanism behind RTFC, the paleomagnetic evidence from Flow 20 certainly strengthens the case that they are happening, not only by representing a second observation of the phenomenon, but one with a remanence acquisition history distinct from the Steens Mountain example.

This latest evidence should re-open the debate about the characteristics of flow in the Earth's outer core and the electrical conductivity of the lower mantle.

"If these kinds of signals are making their way out of the core," Glen posits, "then that tells us something about conductivity in the mantle. Perhaps there are large heterogeneities that give us windows through to the core from which flux lines can make their way out quickly and unperturbed."

The last reversal on record is the Brunhes-Matuyama reversal, named after Bernard Brunhes and Motonori Matuyama. It occurred approximately 780,000 years ago. At present, the overall geomagnetic field has been steadily weakening (declining 10-15% in the last 150 years) since it was first measured, leading to speculation that the Earth is headed for yet another reversal. What would happen during a field reversal is anybody’s guess since humanity has not yet experienced one in the age of modern instrumentation.

Speculating on the effects of rapid field reversal on the Earth's biosphere, Glen says, "There are lots of animals who use the magnetic field for navigation. The majority are bacteria who use it to navigate to find food. They could be significantly affected. They have relatively short life-cycles and occur in large numbers so they can adjust rapidly compared to higher organisms. On the other hand, it appears that many higher organisms have other tools at their disposal to find their way around and are not solely dependent on the magnetic field to navigate."

When asked about the possible effects on Homo sapiens, he says, "If we were to undergo a polarity reversal and experience this today, it would certainly have profound effects on the way we do things. It would significantly affect space weather, which would affect satellites, and communications and power systems. But we have really good records of when these events have occurred in the past, and for the most part, they don't correlate with mass extinctions."

For the most part?

Glen laughs, and then explains, "It's difficult to do the correlation to assess whether there are more subtle effects on life because rarely do you find good magnetic records in the same stratigraphic sections as good biologic records. The largest mass extinction, the Permian-Triassic, as far as the resolution and magnetic record and dating can tell us, coincided with a polarity reversal. Although the likelihood of two very large and rare events like this coinciding is rare, this type of coincidence can, and given enough time, will occur".

Possible apocalypse aside, Glen and his colleague will continue work in Sheep Creek, gathering more data, and furthering understanding within this relatively young branch of geoscientific study.

"As with any research, you never know where it will end up leading," Glen says. "But my main drive, and I suspect why Scott does this too, is sheer awe and curiosity. Here's this really important earth process that we know very little about. The fact that we know so little about it; it's something you can actually contribute to."