Gubbins D., Herrero-Bervera E. Encyclopedia of Geomagnetism and Paleomagnetism

Подождите немного. Документ загружается.

Cowling, T.G., 1934. The magnetic field of sunspots. Monthly Notices

of the Royal Astronomical Society, 94:39–48.

Merrill, R.T., and McFadden, P.L., 1999. Geomagnetic polarity transi-

tions. Reviews of Geophysics, 37: 201–226.

Merrill, R.T., McElhinny, M.W., and McFadden, P.L., 1996. The Mag-

netic Field of the Earth: Paleomagnetism, the Core, and the Deep

Mantle. San Diego, CA: Academic Press.

Cross-references

Cowling’s Theorem

Geocentric Axial Dipole Hypothesis

Geomagnetic Polarity Timescales

Geomagnetic Secular Variation

Harmonics, Spherical

Inner Core Anisotropy

Nondipole Field

Reversals, Theory

Westward Drift

GEOMAGNETIC HAZARDS

Introduction

Geomagnetic variations take place over a wide range of timescales.

The longer-term variations, typically those occurring over decades to

millennia, are predominantly the result of dynamo action in the Earth’s

core (see Geodynamo). However, geomagnetic variations on time-

scales of seconds to many years also occur due to dynamic processes

in the ionosphere (see Ionosphere), magnetosphere (see Magneto-

sphere of the Earth), and heliosphere (see Magnetic field of Sun).

These changes are ultimately tied to variations associated with the

solar activity cycle. To a lesser extent geomagnetic variations are also

caused by changes in the solar ultraviolet emission that controls the

ionization of the Earth’s ionosphere (see Periodic external fields).

However, the largest geomagnetic variations are due to sporadic solar

activity in the form of solar coronal mass ejections (CMEs) and the

number of CMEs varies with the phase of the solar cycle.

In a CME, the rapid reconfiguration, through magnetic reconnection

processes, of the fields in and above solar sunspot regions results in

the release of magnetic flux and plasma into the solar wind. Traveling

toward Earth, these plasma “bubbles” can interact with the Earth’s

magnetic field, depositing particles and energy into the magneto-

sphere, and driving geomagnetic storms (see Storms and substorms,

magnetic).

Solar activity, and hence geomagnetic storm frequency, varies on a

cycle of between about 9 and 14 years (see Dynamo, solar). A large-

scale dipolar solar magnetic field exists during times of sunspot mini-

mum. This dipolar field progressively diminishes through solar (sunspot)

maximum, becoming reestablished with the opposite polarity around

the time of the next cycle minimum. Sunspots, flares, and associated

CMEs are the means by which the large-scale field is changed and

magnetic flux is expelled into the solar wind (see Magnetic field

of Sun).

The fact that the geomagnetic field does respond to solar conditions

can be useful, for example, in investigating Earth structure using mag-

netotellurics (see Magnetotellurics), but it also creates a hazard. This

geomagnetic hazard is a risk to technology, rather than to health (for

recent reviews see, e.g., Lanzerotti, 2001; Pirjola et al., 2005). Astro-

nauts are certainly at risk from bursts of ionizing solar energetic

particle radiation that follow solar flares (Turner, 2001). Astronaut

protection involves appropriate spacecraft shielding and positioning

relative to the Sun. However, at the Earth’s surface, the atmosphere

acts as a protective layer equivalent to several meters of concrete. Only

in the polar regions and at airline altitudes anywhere may solar radia-

tion storms be a health issue, and major airline operators and aviation

authorities are considering the risks (Getley, 2004). In terms of hazard

to health, modeling and predicting the changing morphology of the

geomagnetic field is important in determining where charged particles

may enter into the lower atmosphere (see Main field modeling;

Geomagnetic secular variation).

Conducting networks and geomagnetically

induced currents

A time-varying magnetic field external to the Earth induces a second-

ary magnetic field, internal to the Earth, as a consequence of Faraday’s

law. Associated with time variations in the induced field is an electric

field, which is measurable at the surface of the Earth. The surface elec-

tric field causes electrical currents, known as geomagnetically induced

currents (GIC), to flow in any conducting structure, for example, a

power grid grounded in the Earth. This electric field, measured in V

1

, acts as a voltage source across networks, with the different

grounding points at different electrical potentials.

Examples of conducting networks are electrical power transmission

grids, oil and gas pipelines, undersea communication cables, telephone

and telegraph networks, and railways. GIC are often described as

being quasi-direct current (DC), although the variation frequency of

GIC is governed by the time variation of the electric field. For GIC

to be a hazard to technology, the current has to be of a magnitude

and frequency that makes the equipment susceptible to either immedi-

ate or cumulative damage. The size of the GIC in any network is partly

governed by the electrical resistance of the network, relative to the

resistance of the underlying Earth. The largest external magnetic field

and magnetospheric current variations occur during geomagnetic

storms and it is then that the largest GIC occur. Significant variation

periods are typically from seconds to about an hour, so the induction

process involves the upper mantle and lithosphere (see Magnetotellu-

rics; Geomagnetic deep sounding). Since the largest magnetic field

variations are observed at higher magnetic latitudes, GIC have been

regularly measured in Canadian, Finnish, and Scandinavian power

grids and pipelines since the 1970s. GIC of tens to hundreds of Amps

have been recorded. However, GIC and effects have also been

recorded in countries at midlatitudes during major storms. There may

even be a risk to low latitude nations during a storm sudden com-

mencement (see Storms and substorms, magnetic) because of the high,

short-period, rate of change of the field that occurs everywhere on the

dayside of the Earth.

GIC have been known since the mid-1800s when it was noted that

telegraph systems could run without power during geomagnetic

storms, described at the time as operating by means of the “celestial

battery” (Boteler et al., 1998). However, technological change and

the growth of conducting networks have made the significance of

GIC greater and more pervasive in modern society. We therefore

describe the GIC hazard in detail in two of the best-studied networks,

power grids and pipelines. The technical considerations for undersea

cables, telephone and telegraph networks, and railways are similar.

However, fewer problems are known, or have been reported in the

open literature, about these systems. This suggests that the hazard is

less, or that there are reliable methods for equipment protection.

Modern electrical transmission systems consist of generating plant

interconnected by electrical circuits that operate at fixed transmission

voltages controlled at transformer substations. The grid voltages

employed are largely dependent on the path length between these

substations and 200–700 kV system voltages are common. There is

a trend toward higher voltages and lower line resistances to reduce

transmission losses over longer and longer path lengths. However,

low line resistances produce a situation favorable to the flow of GIC.

Power transformers have a magnetic circuit that is disrupted by the

quasi-DC GIC: the field produced by the GIC offsets the operating point

of the magnetic circuit and the transformer may go into half-cycle

316 GEOMAGNETIC HAZARDS

saturation. This produces a harmonic-rich AC waveform, localized heat-

ing, and leads to high reactive power demands, inefficient power trans-

mission and possible misoperation of protective measures. Balancing

the network in such situations requires significant additional reactive

power capacity (Erinmez et al., 2002). The magnitude of GIC that will

cause significant problems to transformers varies with transformer type.

Modern industry practice is to specify GIC tolerance levels on new

transformer purchases.

On 13 March 1989 a severe geomagnetic storm caused the collapse

of the Hydro-Quebec power grid in a matter of seconds as equipment

protection relays tripped in a cascading sequence of events (Bolduc,

2002). Six million people were left without power for 9 hours, with

significant economic loss. Since 1989 power companies in North

America, the UK, Northern Europe, and elsewhere have invested time

and effort in evaluating the GIC risk and in developing mitigation stra-

tegies. GIC risk can, to some extent, be reduced by capacitor blocking

systems, maintenance schedule changes, additional on-demand generat-

ing capacity, and, ultimately, shedding of load. However, these options

are expensive and sometimes impractical. The continued growth of

high voltage power networks, for example, in North America and in

mainland Europe, is leading to a higher risk. This is partly due to

the increase in the interconnectedness at higher voltages; connections

to grids in the auroral zone, and commercial considerations that see

grids run closer to capacity than was the case historically.

To understand the flow of GIC in power grids and therefore to

advise on GIC risk, analysis of the quasi-DC properties of the grid is

necessary. This must be coupled with a geophysical model of the Earth

that provides the driving surface electric field, determined by combin-

ing time-varying ionospheric source fields and a conductivity model of

the Earth. Such analyses have been performed for North America, the

UK, and in Northern Europe. However, the complexity of power grids,

the source ionospheric current systems, and the 3D ground conductiv-

ity makes an accurate analysis difficult. By being able to analyze,

postevent, major storms, and their consequences we can build a picture

of the weak spots in a given transmission system and even run

hypothetical event scenarios. An example of a postevent grid analysis

for Central Scotland and Northern England during the peak of the 30

October 2003 magnetic storm is shown in Figure G20 (Thomson

et al., 2005). At this time the measured GIC was 42 A, a significant,

though in this case, manageable level and comparable with the model

results.

Grid management is also aided by space weather forecasts of major

geomagnetic storms. This allows for mitigation strategies to be imple-

mented. Solar observations provide a 1–3 day warning of an Earth-

bound CME, depending on CME speed. Following this, detection of

the solar wind shock that precedes the CME in the solar wind, by

spacecraft at the Lagrangian L1 point, gives a definite 20–60 min

warning of a geomagnetic storm (again depending on local solar wind

speed). However, the magnitude and accurate time of arrival of a CME

prior to shock detection is unknown, although there is much research

and model development within the space weather community. Given

the demands on managing power grids more accurate information

would have high value.

Major pipeline networks exist at all latitudes and many systems are

on a continental scale. Pipeline networks are constructed from steel to

contain high-pressure liquid or gas and are covered with special coat-

ings to resist corrosion. Weathering and other damage to the pipeline

coating can result in the steel being exposed to moist air or to the

ground, causing localized corrosion problems. Cathodic protection

rectifiers are used to maintain pipelines at a negative potential with

respect to the ground. This minimizes corrosion without allowing

any chemical decomposition of the pipe coating and the operating

potential is determined from the electrochemical properties of the soil

and Earth in the vicinity of the pipeline. The GIC hazard to pipelines is

that GIC cause swings in the pipe-to-soil potential, increasing the rate

of corrosion during major geomagnetic storms (Gummow, 2002).

GIC risk is not, therefore, a risk of catastrophic failure, rather the

reduced service lifetime of the pipeline, or parts of it.

Pipeline networks are modeled in a similar manner to power grids,

for example, through distributed source transmission line models that

provide the pipe-to-soil potential at any point along the pipe (Boteler,

1997). These models need to take into account complicated pipeline

topologies that include bends as well as electrical insulators, or

flanges, that electrically isolate different sections of the network. From

a detailed knowledge of the pipeline response to GIC, pipeline engi-

neers can understand the behavior of the cathodic protection system

even during a geomagnetic storm, when pipeline surveying and main-

tenance may often be suspended.

Satellite operation, navigation, and radio

communication

The ionosphere plays a significant role in very low frequency (VLF)

through to high-frequency (HF) radio communication, and in naviga-

tion systems such as Loran-C and Omega (Lanzerotti, 2001; Schunk,

2001). Ionospheric conductivity is partly affected by geomagnetic

storms but more significantly by solar ultraviolet and x-ray control

of the ionospheric D, E, and F layers (see Ionosphere). Solar flares

cause signal-phase anomalies and amplitude variations to occur (fades

and enhancements) and conditions can persist for minutes to hours.

Solar flares and CMEs are also important as sources of solar energetic

particles, affecting radio communications at high latitudes. Disturbed

conditions can persist for days to weeks in the Polar Regions during

periods of high particle flux into the polar caps. Ultrahigh-frequency

(UHF) radio signals are central to the Global Positioning System

(GPS) that utilizes satellites in Earth orbit for precise ground position

determination. UHF waves pass largely unattenuated through the

Figure G20 Estimated GIC in a network model of the Central

Scotland and Northern England high voltage power grid, at the

peak of the 30 October 2003, storm at 21:20 UT. Circle shading

denotes GIC flowing to/from earth (dark/light) at transformers

in the grid and the arrows denote the instantaneous E and B fields

at Eskdalemuir magnetic observatory. The E-field is calculated

from a radially varying Earth conductivity model using

measured Eskdalemuir geomagnetic variations at this time.

GIC amplitudes are proportional to spot size, with 40 A shown for

scale. White spots show the locations of permanent GIC

measurement sites in the grid.

GEOMAGNETIC HAZARDS 317

ionosphere but the system accuracy is sensitive to variations in the

total electron content (TEC) in the path between ground and satellite.

The TEC determines the signal propagation delay. TEC variations

occur during geomagnetic storms and these particularly degrade the

accuracy of single-frequency GPS equipment. Geomagnetic storms

also produce ionospheric irregularities and scintillations that occur

both on the dayside and nightside of the Earth.

In practical terms, radio navigational and communication systems

and operators adjust to the prevailing conditions, although accurate

forewarning of solar and geomagnetic activity may be useful for plan-

ning purposes.

Low Earth orbit satellites and space stations (up to around 1000 km

altitude) experience increased air drag during geomagnetic storms.

Enhanced ionospheric currents deposit heat in the atmosphere. This

causes the atmosphere to expand outward and, at a given altitude,

atmospheric density increases. This leads to heightened drag forces,

slowing of satellite velocities, and lowering of orbit altitudes. The

cumulative effect of geomagnetic storms is therefore to reduce the

operational lifetime of satellites, particularly those in low initial orbits,

where air density is higher. High-altitude atmospheric density models

often parameterize geomagnetic heating effects by geomagnetic activ-

ity indices (e.g., Roble, 2001). By predicting geomagnetic indices, on

day-to-day and solar cycle timescales, more efficient use can be made

of satellite fuel supplies, with judicious orbit reboosts used to extend

mission lifetimes. More sophisticated models of the upper atmosphere

are being developed, involving near real-time calibration data from

many orbiting satellites, as well as a better understanding and model-

ing of the physics of the atmosphere and its response to solar and

geomagnetic forcing.

Satellites also suffer increased surface and internal electrical char-

ging from ionized particles during geomagnetic storms. This can result

in system malfunctions as electronic components experience physical

damage and logic errors. Satellite manufactures take the solar cycle

varying radiation environment into account when designing compo-

nents, using statistical models of radiation dosages over component

lifetimes. Modeling of the geomagnetic field morphology and predict-

ing changes in the field (see Main field modeling; Geomagnetic secu-

lar variation) help to map the radiation environment of the Earth and

the response of satellites to that environment. In operation, satellites

can be temporarily powered down or placed in an appropriate “safe

mode” following warnings of solar and geomagnetic activity.

Geophysical exploration and geomagnetic variations

Aeromagnetic surveys are affected by geomagnetic storm variations

(see Crustal magnetic field; Magnetic anomalies for geology and

resources; Aeromagnetic surveying). These cause data interpretation

problems where external field amplitudes are similar to those of the

crustal field in the survey area. Accurate geomagnetic storm warnings,

including an assessment of the magnitude and duration of the storm,

would allow for an economic use of survey equipment.

For economic and other reasons, oil and gas exploration often

involves the directional drilling of well paths many kilometers from

a single wellhead in both the horizontal and vertical directions. Target

reservoirs may only be a few tens to hundreds of meters across and

accurate surveying by gyroscopic methods is expensive since it can

involve the cessation of drilling for a number of hours. An alternative

is to use magnetic referencing while drilling (Clark and Clarke, 2001).

Near real-time magnetic data are used to correct the drilling direction

and nearby magnetic observatories prove vital. There is no drilling

“down-time” during a magnetic storm and storm forecasts are not

normally seen as being important.

Summary and outlook

The geomagnetic hazard to technology results from the strengthening

of magnetospheric and ionospheric current systems by the solar wind

and by CMEs. These electrical current enhancements cause rapid and

high amplitude magnetic variations during geomagnetic storms. Pro-

cesses internal to the magnetosphere also drive variations, particularly

after CME-driven activity. A common theme that emerges from a

study of geomagnetic hazards is a need for accurate geomagnetic

storm forecasting, in terms of onset time and duration, maximum

amplitude, and variation periods. The close connection of geomagnetic

hazard with solar activity and space weather is also clear. Some prac-

tical applications of geomagnetic variations require only monitoring

data, but other applications (will) clearly benefit from a thorough phy-

sical understanding of the Sun-Earth magnetic interaction and, in

particular, accurate prediction of geomagnetic variations.

Alan W.P. Thomson

Bibliography

Bolduc, L., 2002. GIC observations and studies in the Hydro-Quebec

power system. Journal of Atmospheric and Solar-Terrestrial

Physics, 64(16): 1793–1802.

Boteler, D.H., 1997. Distributed source transmission line theory for

electromagnetic induction studies. In Supplement of the Proceed-

ings of the 12th International Zurich Symposium and Technical

Exhibition on Electromagnetic Compatibility, February 18 to 20,

1997 at the Swiss Federal Institute of Technology in Zurich,

Switzerland, pp. 401–408.

Boteler, D.H., Pirjola, R.J., and Nevanlinna, H., 1998. The effects of

geomagnetic disturbances on electrical systems at the Earth’s sur-

face. Advances in Space Research, 22(1): 17–27.

Clark, T.D.G., and Clarke, E., 2001. In Space Weather Workshop:

Space weather services for the offshore drilling industry. Looking

Towards a Future European Space Weather Programme. ESTEC,

ESA WPP-194.

Erinmez, I.A., Kappenman, J.G., and Radasky, W.A., 2002. Manage-

ment of the geomagnetically induced current risks on the national

grid company’s electric power transmission system. Journal of

Atmospheric and Solar-Terrestrial Physics, 64(5–6): 743–756.

Getley, I.L., 2004. Observation of solar particle event on board a com-

mercial flight from Los Angeles to New York on October 29, 2003.

AGU Space Weather, 2: S05002, doi:10.1029/2003SW000058.

Gummow, R.A., 2002. GIC effects on pipeline corrosion and corro-

sion-control systems. Journal of Atmospheric and Solar-Terrestrial

Physics, 64(16): 1755–1764.

Lanzerotti, L.J., 2001. Space weather effects on technologies. In Song,

P., Singer, H.J., and Siscoe, G.L. (eds.) Space Weather. Geophysi-

cal Monograph 125. Washington, DC: American Geophysical

Union, pp. 11–22.

Pirjola, R., Kauristie, K., Lappalainen, H., and Viljanen, A., 2005.

Space weather risk. AGU Space Weather, 3: S02A02, doi:10.1029/

2004SW000112.

Roble, R.G., 2001. On forecasting thermospheric and ionospheric

disturbances in space weather events. In Song, P., Singer, H.J.,

and Siscoe, G.L. (eds.) Space Weather. Geophysical Monograph

125. Washington, DC: American Geophysical Union, pp. 369–376.

Schunk, R.W., 2001. Ionospheric climatology and weather distur-

bances: a tutorial. In Song, P., Singer, H.J., and Siscoe, G.L.

(eds.) Space Weather. Geophysical Monograph 125. Washington,

DC: American Geophysical Union, pp. 359–368.

Thomson, A.W.P., McKay, A.J., Clarke, E., and Reay, S.J., 2005.

Surface electric fields and geomagnetically induced currents in

the Scottish power grid during the October 30, 2003, geomagnetic

storm. AGU Space Weather, 3: S11002, doi:10.1029/2005SW

000156.

Turner, R., 2001. What we must know about solar particle events

to reduce the risk to astronauts. In Song, P., Singer, H.J., and

Siscoe, G.L. (eds.) Space Weather.

Geophysical Monograph 125.

Washington, DC: American Geophysical Union, pp. 39–44.

318 GEOMAGNETIC HAZARDS

Cross-references

Aeromagnetic Surveying

Crustal Magnetic Field

Dynamo, Solar

Geodynamo

Geomagnetic Deep Sounding

Geomagnetic Secular Variation

Ionosphere

Magnetic Anomalies for Geology and Resources

Magnetic Field of Sun

Magnetosphere of the Earth

Magnetotellurics

Main Field Modeling

Periodic External Fields

Storms and Substorms, Magnetic

GEOMAGNETIC JERKS

Observations

Geomagnetic jerks are abrupt changes in the second-time derivative, or

secular acceleration, of the magnetic field that arises from sources

inside the Earth. They delineate intervals of oppositely signed, near-

constant secular acceleration. The first observed geomagnetic jerk

was that around 1969 (Courtillot et al., 1978), and since the late

19th century when direct and continuous measurements of the Earth’s

magnetic field have been made at a number of observatories around

the world, geomagnetic jerks have also been observed to occur around

1925, 1978, 1991, and 1999 and possibly also around 1901 and 1913

(e.g., Malin and Hodder, 1982; Courtillot and Le Mouël, 1984;

Macmillan, 1996; Alexandrescu et al., 1996; Mandea et al., 2000).

These jerks are most readily observed in the first-time derivative of the

east component at European observatories (Figure G21). Understanding

their origin is important, not only because they result from interesting

dynamical processes in the core and may help determine the conductiv-

ity of the mantle, but also for improving time-dependent models of

the geomagnetic field (q.v.) and for the strictly practical purpose of

forecasting its future behavior, for example, as used in navigation.

Analysis of jerks

Many studies have been undertaken to assess jerk characteristics, such

as origin, times of occurrence, and spatial patterns. At observatories

the measured magnetic field is the vector sum of fields arising from

three primary sources, namely the main field generated in the Earth’s

core, the crustal field from local rocks, and a combined disturbance

field from electrical currents flowing in the upper atmosphere and

magnetosphere, which also induce electrical currents in the sea and

the ground. As the long-period variations of the disturbance field asso-

ciated with the 11-year solar cycle are in the same frequency band

as variations arising from sources inside the Earth, the separation of

the sources is an important part of any analysis of geomagnetic

jerks. Spherical harmonic analysis applied to monthly or yearly mean

geomagnetic observatory data has shown that jerks are internal in ori-

gin and, where sufficient data exist, are global phenomena (e.g., Malin

and Hodder, 1982, McLeod, 1985).

Various analysis techniques have been applied to jerks to investigate

specific aspects of their temporal and spatial characteristics. Methods

such as optimal piecewise regression analysis and wavelet analysis

have established the exact dates of occurrence at different observatories

without any a priori assumptions. Using wavelets, the 1969 and 1978

jerks have been shown to have different arrival times at the Earth’s

surface, with the northern hemisphere leading the southern hemisphere

by about 2 years (Alexandrescu et al., 1996).

Implications for studies of the Earth’s deep interior

The shortest variations in the Earth’s internal field that can be seen at

the Earth’s surface are determined by the rate of change of the mag-

netic field at the core-mantle boundary (CMB) and by the electrical

conductivity of the mantle through which the magnetic signal from

the core must pass. From analyses of surface data we know that the

duration of individual jerks, as indicated by the width of impulses in

the third-time derivative, is no more than about 2 years. However, it

Figure G21 Geomagnetic jerks as seen in the secular variation of the east component of the Earth’s magnetic field observed at

European observatories. Time of jerks are shown by arrows.

GEOMAGNETIC JERKS 319

is difficult to make any estimate of the rate of change of the magnetic

field at the CMB, or the electrical conductivity of the mantle, from

geomagnetic field observations at the Earth’s surface. As in most geo-

physical problems, assumptions have to be made or constraints have to

be applied using information from other data sources. In the case of

geomagnetic jerks one useful independent dataset is length of day

(LOD) observations. Correlation of the occurrences of jerks with

marked changes in the deficit LOD (the former appearing to lead the

latter by a few years) indicates some form of coupling between the

core and the mantle. It has recently been shown that inflexion in the

time derivative of splines fitted to filtered LOD data from 1960

onwards coincide remarkably well with geomagnetic jerks, including

the late arrivals reported for the southern hemisphere (Holme and

de Viron, 2005).

Jerks are most likely to result from dynamic processes inside the

Earth that produce changes in the magnetic field near the top of the

fluid core. These include the changes caused by advection and shear

within the flow. Diffusion is not likely to be an important factor

because of the short timescales involved. Core flows can be estimated

directly from observatory secular variation data or indirectly from

spherical harmonic models of secular variation. Again, assumptions

have to be made or constraints applied so that a unique solution is

found. One common assumption applied when determining core flows

from magnetic data over time intervals of a few decades is the frozen-

flux hypothesis with tangentially geostrophic time-varying flows. With

this assumption the fine detail of the secular variation, including jerks,

can be well fitted (Jackson, 1997) but the models offer little physical

explanation of the flows in terms of core processes. Another assump-

tion that also gives good fits to data is steady flow in a core reference

frame that is allowed to rotate about the Earth’s rotation axis with

respect to the mantle (Holme and Whaler, 2001). This, not surpris-

ingly, provides results which are consistent with LOD data.

A time-dependent flow model that comprises only torsional oscilla-

tions, recovers jerks very well while not producing spurious jerks

where none are observed (Bloxham et al., 2002). This suggests that

the origin of jerks is related to the action of torsional oscillations on

the local magnetic field at the CMB. If this is correct, then the tor-

sional oscillations provide a physical representation for the origin of

jerks, for they are not only a part of the mathematical representation

of core flows but, more importantly, are an actual component of flow

predicted to occur in the geodynamo. They also provide the angular

momentum changes required to explain the LOD changes. However,

the specific details of jerk generation in the core are still to be

resolved, in particular the mechanism driving torsional oscillations.

Susan Macmillan

Bibliography

Alexandrescu, M., Gibert, D., Hulot, G., Le Mouël, J.-L., and Saracco,

G., 1996. Worldwide wavelet analysis of geomagnetic jerks.

Journal of Geophysical Research, 101(B10): 21975–21994.

Bloxham, J., Zatman, S., and Dumberry, M., 2002. The origin of

geomagnetic jerks. Nature, 420:65–68.

Courtillot, V., Ducruix, J., and Le Mouël, J.-L., 1978. Sure une accél-

eration récente de la variation seculaire du champ magnétique

terrestre. Comptes rendus des séances de l’ Académie des Sciences,

287, D: 1095–1098.

Courtillot, V., and Le Mouël, J.-L., 1984. Geomagnetic secular

variation impulses. Nature, 311: 709–716.

Holme, R., and de Viron, O., 2005. Geomagnetic jerks and a high-reso-

lution length-of-day profile for core studies. Geophysical Journal

International, 160: 413–413. doi: 10.1111/j.1365-

246X.2005.02547.x.

Holme, R., and Whaler, K.A., 2001. Steady core flow in an azimuth-

ally drifting reference frame. Geophysical Journal International,

145: 560–569.

Jackson, A., 1997. Time-dependency of tangentially geostrophic core

surface motions. Physics of the Earth and Planetary Interiors,

103: 293–312.

Macmillan, S., 1996. A geomagnetic jerk for the early 1990’s. Earth

and Planetary Science Letters, 137: 189–192.

Malin, S.R.C., and Hodder, B.M., 1982. Was the 1970 geomagnetic

jerk of internal or external origin? Nature, 296: 726–728.

Mandea, M., Bellanger, E., and Le Mouël, J.-L., 2000. A geomagnetic

jerk for the end of the 20th century? Earth and Planetary Science

Letters, 183: 369–373.

McLeod, M.G., 1985. On the geomagnetic jerk of 1969. Journal of

Geophysical Research, 90(B6): 4597–4610.

Cross-references

Decade Variations in LOD

Main Field Modelling

Spherical Harmonics

Time-dependent Models of the Main Magnetic Field

Torsional Oscillations

GEOMAGNETIC POLARITY REVERSALS

Early records of reversals

Natural magnetization of rocks opposed to that of the present day field

was observed by David and Brunhes about one century ago. Brunhes

concluded that “at a certain moment of the Miocene epoch in the

neighborhood of Saint-Flour the north pole was directed upward: it

was the south pole which was the closest to central France” (see Laj

et al., 2002). Then, Matuyama observed in 1929 other reverse polarities

in the magnetic directions of lava flows from Japan and Manchuria.

However, early studies did not provide definite evidence that the geomag-

netic field had reversed its polarity, in particular because self-reversal in

some rocks can produce thermoremanent magnetization antiparallel to

the geomagnetic field at cooling time. Existence of polarity reversals

was definitely established when K-Ar dating demonstrated that all lavas

from the same age have a similar polarity, whatever their geographical

position at the Earth surface. First Geomagnetic Polarity Timescales

(GTPS) were constructed in the 1960s (Cox, 1969). Then, discovery of

marine magnetic anomalies (Figure G22) confirmed seafloor spreading

(Vine and Matthews, 1963), and the GTPS was extended to older times

(Vine, 1966; Heirtzler et al., 1968; Lowrie and Kent, 1981). Since then,

succession of polarity intervals has been extensively studied and used

to construct magnetostratigraphic timescales linking biostratigraphies,

isotope stratigraphies, and absolute ages (see Opdyke and Channell,

1996, for a review).

Reversal frequency

The geomagnetic polarity timescale constructed from seafloor mag-

netic anomalies (Figure G23) revealed that polarity reversals were

common in Earth’s history. Their occurrence, however, is not constant

through time. Reversal frequency has increased during the past 85 Ma,

since the long Cretaceous interval of normal polarity. Before this

superchron, reversal frequency decreased since the Late Jurassic.

Changes in reversal frequency occur with time constant of tens of

millions of years or more, of the same order of magnitude of those

involved in mantle overturning. It was therefore suggested that rever-

sal frequency is linked to the structure of the mantle, an idea that

was reinforced by 3D computations of the geodynamo (Glatzmaier

et al., 1999). The question of the existence of a 15 Ma periodicity in

reversal frequency (Mazaud et al., 1983) has been largely debated

(see Jacobs, 1994).

320 GEOMAGNETIC POLARITY REVERSALS

Geometry and dynamics of the transitional field

Polarity transitions occur so quickly on a geological timescale that it is

difficult to find rocks that have preserved in detail variations of the

transitional field. Also, sediment and lava may sometimes provide

different views of the phenomenon. Lava flows acquire magnetic

memory during their cooling, and provide instantaneous pictures of the

Earth magnetic field. Polarity transitions are recorded only if eruptions

occurred at the time of the transition, and if no alteration or secondary

heating affected the recording rocks. Lava may provide absolute

paleointensities, by using Thellier, Shaw, or Thellier and Coe determi-

nation methods. In contrast, sediments provide continuous records of

geomagnetic field changes, but acquisition process may smooth out fast-

est field changes. Only relative changes of past geomagnetic intensity

are obtained, and only with sediment exhibiting limited downcore

variations in magnetic mineralogy and magnetic particle grain size.

Despite difficulties to obtain paleomagnetic records of polarity tran-

sitions, several characteristics have emerged in the last few decades.

The geomagnetic intensity decreases a few thousands of years before

directional changes. Then, the magnetic vector exhibit directional

changes while intensity remains low. When the direction reaches oppo-

site polarity, the geomagnetic intensity rises to normal values. The

total process takes place in few thousand years, and may vary for dif-

ferent reversals. It was also suggested that the geomagnetic field inten-

sity progressively decreases during periods of stable polarity and

recovers high values immediately after a transitio n (Valet and Meynadier ,

1993).

The standing field hypothesis and the flooding models

The geometry and the dynamics of the geomagnetic field during polar-

ity transitions have been subject to large debate. Hillhouse and Cox

(1976) suggested that if the usual nondipole drift (secular variation)

was unchanged during a polarity reversal, which is theoretically envi-

saged (Le Mouël, 1984), then one should observe large longitudinal

swings in reversal records. Their absence in the first records obtained

from sediments led these authors to suggest the standing field hypoth-

esis, in which an invariant component dominates the transitional field

when the axial dipole component has vanished. The standing field

hypothesis, however, was not confirmed. A classical tool used for

investigating the morphology of the transitional field is the virtual geo-

magnetic pole (VGP), defined as the pole of the dipolar field that gives

the observed direction of magnetization at studied site. If VGPs

obtained at several sites for a given instant in time coincide, then the

field has a dipolar structure. If different VGPs are obtained at different

sites, then the field was not dipolar (the word “virtual” indicates

that VGPs can be calculated for any field, not necessarily dipolar).

Figure G23 Geomagnetic polarity timescale with magnetic

anomaly numbers for the past 160 Ma. (Figure from Merrill et al.,

1996).

Figure G22 Magnetic structure of seafloor near the Reyjkanes Ridges (after Phinney, 1968).

GEOMAGNETIC POLARITY REVERSALS 321

Sedimentary records also indicated different VGP paths for the same

transition studied at different sites at the Earth’s surface, with transi-

tional VGPs moving progressively along longitudinal great circles.

This led Hoffman and Fuller (1978) to develop flooding models, in

which reversals originate in a localized region of the core and then

progressively propagate into other regions.

Stop and go behavior

Higher resolution records were obtained from lava and sedimentary

series. They suggested some repetition over successive reversals (Valet

and Laj, 1984), and also stop-and-go behavior, with alternating phases of

rapid change and stationarity during transitions. This is seen in the Mio-

cene reversal record obtained at Steens Mountain, Oregon (Figure G24)

(Prévot et al., 1985), and also in Miocene sedimentary records from

Northwest Greece (Laj et al., 1988).

The two preferred bands

An intriguing point was issued by Laj and colleagues (1991), examin-

ing the longitudinal distribution of the VGP paths of sedimentary

records of reversals over the past 10 Ma. They found that VGP paths

tend to follow paths located into two preferred bands, over Americas

and eastern Asia. These two preferred bands coincide with regions of

fast seismic wave propagation in the lower mantle, in the downward

continuation of subducted slabs in the mantle (Figure G25). These

two bands also coincide with the main patches of radial magnetic flux

at the core-mantle boundary in the present-day field and the historical

field (Bloxham and Gubbins, 1985), and in paleofield reconstructions

(Constable, 1992; Gubbins and Kelly, 1993). Thus, Laj and colleagues

(1991) argued that the persistence of these two preferred bands

over time of the order of magnitude of mantle convection was experi-

mental evidence that the geodynamo was constrained by temperature

patterns at the core-mantle boundary, as suggested by present day

Figure G24 The Steens Mountain directional record. Stereographic projection of field directions after rotation about the east-west

horizontal axis, so dipole field directions coincide with the pole of the projection sphere (from Pre

vot et al., 1985).

322 GEOMAGNETIC POLARITY REVERSALS

and historical magnetic observations. The existence of the two pre-

ferred bands was and is still debated, in particular because they were

not seen in the volcanic compilation of Prévot and Camps (1993).

However, another analysis of volcanic databases suggests that the

two bands also exist in compilations of volcanic reversals records

(Love, 1998). Influence of variable conditions at the core-mantle

boundary is suggested in 3D computations of the geodynamo (Glatz-

maier et al., 1999).

Toward a consistent picture of polarity reversals?

Progressively, a consistent picture emerges from high-resolution sedi-

ment and lava records. In 1992, Hoffman observed long-lived transi-

tional states of the geomagnetic field, with VGPs of volcanic

reversal records clustering in the two preferred bands (Figure G26).

More recently, several high-resolution sedimentary records were

obtained that exhibited complex field behavior, with VGP loops and

clusters reminiscent of volcanic records. Some clusters, but not all,

lie in the two preferred bands (Figure G27). Transitional precursors

are sometimes observed.

Conclusion

The situation is still complex, because different reversals may docu-

ment different behavior and the question of the two preferred longitu-

dinal bands is still open. The relation between polarity reversals and

excursions is under investigation. Geomagnetic excursions are seen

as aborted reversals in which the field may reverse in the liquid outer

core, which has timescales of 500 years or less, but not in the solid

inner core, where field must change by diffusion with a timescale of

3 ka. This disparity of dynamical timescales between the inner and

outer core is consistent with the presence of several excursions

between full reversals (Gubbins, 1999). Overall, the mechanisms that

trigger reversals and excursions have to be better understood. Whether

or not the present-day axial dipole field decrease corresponds to a

reversal or excursion onset, or to a field fluctuation during stable polar-

ity, is not yet known.

Combination of supercomputer 3D simulations (see for instance

Glatzmaier and Olson, 2005, for a review) and new high-resolution

records from both sediments and lava flows should lead to a better

understanding of the mechanisms involved in field generation and

Figure G25 The two preferred bands for VGP paths of sedimentary records of reversals (from Laj et al., 1991). Zones of fast seismic

velocity in the lower mantle are indicated in dark (blue in online version of encyclopedy).

Figure G26 Clusters of VGPs (from Hoffman, 1992).

GEOMAGNETIC POLARITY REVERSALS 323

polarity transitions, and ultimately of the global dynamics of the

Earth’s interior. LSCE contribution n



2407.

Alain Mazaud

Bibliography

Bloxham, J., and Gubbins, D., 1985. The secular variation of the

Earth’s magnetic field. Nature, 317: 777–781.

Channell, J.E.T., and Lehman, B., 1997. The last two geomagnetic

polarity reversals recorded in high-deposition-rate sediment drifts.

Nature, 389: 712–715.

Constable, C., 1992. Link between geomagnetic reversal paths and

secular variation of the field over the past 5 Ma. Nature, 358:

230–233.

Cox, A., 1969. Geomagnetic reversals. Science, 263: 237–245.

Glatzmaier, G.A., and Olson, P., 2005. Probing the geodynamo. Scien-

tific American, 292:51–57.

Glatzmaier, G.A., Coe, R., Hongre, L., and Roberts, P.H., 1999. The

role of the Earth’s mantle in controlling the frequency of geomag-

netic reversals. Nature, 401: 885–890.

Gubbins, D., 1999. The distinction between geomagnetic excursions

and reversals. Geophysical Journal International, 137:F1–F3.

Gubbins, D., and Kelly, P., 1993. Persistent patterns in the geomag-

netic field over the past 2.5 Ma. Nature, 365: 829–832.

Heirtzler, J.R., Dickson, G.O., Herron, E.M., Pitman W.C. II, and Le

Pichon, X., 1968. Marine magnetic anomalies, geomagnetic field

reversals, and motions of the ocean floor and continents. Journal

of Geophysical Research, 73:2119–2136.

Hillhouse, J.W., and Cox, A., 1976. Brunhes-Matuyama polarity tran-

sition. Earth and Planetary Science Letters, 29:51–64.

Hoffman, K.A., and Fuller, M., 1978. Polarity transition records and

the geomagnetic dynamo. Nature, 273: 715–718.

Hoffman, K.A., 1992. Dipolar reversal states of the geomagnetic field

and core mantle dynamics. Nature, 359: 789–

794.

Jacobs, J.A., 1994. Reversals of the Earth’s Magnetic Field, 2nd edn.

Cambridge: Cambridge University Press, 339 pp.

Laj, C., Guitton, S., Kissel, C., and Mazaud, A., 1988. Complex beha-

vior of the geomagnetic field during three successive polarity

reversals, 11–12 Ma BP. Journal of Geophysical Research, 93:

11655–11666.

Laj, C., Mazaud, A., Weeks, R., Fuller, M., and Herrero-Bervera, H.,

1991. Geomagnetic reversals paths. Nature, 351: 447.

Laj, C., Kissel, C., and Guillou, H., 2002. Brunhes research revisited:

magnetization of volcanic flows and backed clays. EOS Transac-

tions: American Geophysical Union, 83(35): 381, 386–387.

le Mouël, J.L., 1984. Outer core geostrophic flow and secular variation

of Earth’s magnetic field. Nature, 311: 734–735.

Love, J.J., 1998. Paleomagnetic volcanic date and geometric regularity

of reversals and excursions. Journal of Geophysical Research, 103:

12435–12452.

Lowrie, W., and Kent, D.V., 1981. One hundred million years of geo-

magnetic polarity history. Geology, 9: 392–397.

Mazaud, A., Laj, C., De Seze, L., and Verosub, K.L., 1983. 15 Ma per-

iodicity in the frequency of geomagnetic reversals since 100 Ma

Reply to McFadden. Nature, 304: 328–330.

Merrill, R.T., McElhinny, M.W., and McFadden, P.L., 1996. The Mag-

netic Field of the Earth. International Geophysics Series 63. San

Diego, CA: Academic Press, 527 pp.

Opdyke, N., and Channell, J.E.T, 1996. Magnetic Stratigraphy. Inter-

national Geophysics Series 64. San Deigo, CA: Academic Press,

346 pp.

Phinney, R.A., 1968. The History of the Earth’s Crust. Princeton, NJ:

Princeton University Press, 244 pp.

Prévot, M., and Camps, P., 1993. Absence of preferred longitudinal

sectors from pole from volcanic records of geomagnetic reversals.

Nature, 366:53–57.

Prévot, M., Mankinen, E.A., Grommé, C.S., and Coe, R., 1985. How

the geomagnetic field vector reverses polarity. Nature, 316:230–234.

Valet, J.P., and Laj, C., 1984. Invariant and changing transitional field

in a sequence of geomagnetic reversals.

Nature, 311: 552–555.

Valet, J.P., and Meynadier, L., 1993. Geomagnetic intensity and rever-

sals during the past four million years. Nature, 366: 234–238.

Vine, F.J., 1966. Spreading of the ocean floor: new evidence. Science,

154: 1405.

Vine, F.J., and Matthews, D.H., 1963. Magnetic anomalies over ocea-

nic ridges. Nature, 199: 947–949.

Cross-references

Paleointensity, Absolute, Determination

Polarity Transition, Paleomagnetic Record

GEOMAGNETIC POLARITY REVERSALS,

OBSERVATIONS

Earth’s magnetic field exhibits the remarkable property of undergoing

a 180



change in directions at geologically frequent, but irregular, inter-

vals. As a result, a compass needle that had been pointing to one geo-

graphic pole (such as North), would point to the opposite geographic

pole. Why the field reverses polarity remains a mystery and solving this

mystery is important for understanding the processes in Earth’s fluid

outer core responsible for generating the geomagnetic field.

The best record of how frequently the geomagnetic field reverses po-

larity comes from marine magnetic anomalies (Cande and Kent, 1995).

Neither the marine anomaly record nor long stratigraphic sequences of

polarity history, often record field directi ons that fall in between the

two stable polarity directions. Given the seafloor spreading and sedi-

mentation rates, this indicates that polarity reversals are rapid events,

Figure G27 Examples of VGP paths for the Bunhes-Matuyama (a) and upper Jaramillo (b) transitions obtained from high deposition rate

sediments in the North Atlantic (after Channell and Lehman, 1997).

324 GEOMAGNETIC POLARITY REVERSALS, OBSERVATIONS

at least on geological timescales. For example the marine magnetic

anomaly record limits the time for a reversal to occur to less than 20 ka.

The first step in attempting to solve the mystery of why Earth’s

magnetic field reverses is to determine what happens as the field

reverses. Paleomagnetic records of polarity transitions are the only

source of information about how the geomagnetic field behaves during

a reversal. Because polarity reversals are very short events, polarity

transition records are difficult to obtain. Not only is it necessary to find

a paleomagnetic recorder that provides great enough time resolution to

catch the field in the act of reversing, but also it is necessary for that

recorder to provide a high fidelity recording of the field, even when

the field was weak.

Two very different types of paleomagnetic recorders have repeatedly

yielded polarity transition records. Volcanic sequences created by rapid

extrusion periods that coincide with a polarity reversal, record the

changing field as successive lava flows cool and become permanently

magnetized in the transitional field. Sequences of sedimentary rocks or

unconsolidated sediments have also proven useful in transition studies.

If sedimentation rates are rapid and continuous enough, a record of a

polarity transition may result as layers of sediment become magnetized

as they accumulate during the reversal.

These two types of paleomagnetic recorders document transitional

fields in fundamentally different ways. Most importantly, the way each

type of recorder becomes magnetized is very different, and these dif-

ferences must be considered carefully when interpreting each type of

record. A second important difference is the two kinds of recorders

provide different types of temporal information about reversing fields.

The processes by which lavas become magnetized are well founded

in the theory of thermoremanent magnetization (TRM). This theory

explains how grains of magnetic minerals become permanently mag-

netized as they cool through their magnetic blocking temperatures in

the presence of an external field. Because the cooling times for typical

lava flows are rapid, it may be assumed that the lava flows provide a

spot reading in time of the magnetic field direction and intensity at

the time of cooling. TRM acquisition theory also provides a basis for

obtaining measures of the strength or intensity of the field, known as

absolute paleointensities.

On the other hand, the processes by which sediments become mag-

netized are less well understood. Although it is clear that many types

of sediments can record the geomagnetic field accurately, as evidenced

by recording the full polarity directions predicted for their site loca-

tion, it is not clear how these magnetizations are acquired. A number

of studies indicate that a postdepositional remanent magnetization pro-

cess that locks-in the magnetization at some depth beneath the seafloor

magnetizes marine sediments. Of particular interest in polarity transi-

tion studies is not so much the depth offset of remanence acquisition,

but rather over what depth range the acquisition process occurs simul-

taneously. If the remanence becomes fixed over a relatively thick depth

interval, the sediment in that layer will average out the changes in the

field that occur. The resulting magnetization therefore would be an

integration of the field during that time, and would provide an average

of those changes in the field. As the thickness of sediment over which

the remanence becomes blocked in becomes thinner, the magnetization

approaches more of a spot reading of the changing field. Unfortu-

nately, the remanence lock-in thicknesses in sediments are not known,

and this presents an important difficulty in interpreting sediment

records of polarity transitions.

Unlike the theory of TRM, theories of how sediments become mag-

netized do not provide a basis for determining absolute field intensi-

ties. But, if conditions are favorable, it is possible to obtain records

of relative changes in field strength (Tauxe, 1993). In other words,

sediments are capable of providing information on the increasing or

decreasing field strength, but as of yet, it is not possible to calibrate

those relative changes to absolute magnitudes of field strength.

In addition to the different ways these two types of recorders

become magnetized, there is also a fundamental difference in the tem-

poral resolution provided by each type of recorder. On the timescales

over which polarity reversals occur, it is not possible to determine

the rates at which lavas are extruded from a volcano, or how much

time has elapsed between successive lava flows. This means, that

while lavas may provide high fidelity recordings of the field, the tim-

ing of the sequence of magnetizations of the flows is unknown. Recent

advances in radiometric dating techniques using

Ar/

Ar and K/Ar

methods (Singer and Pringle, 1996) are approaching the precision

needed to provide bounding limits on the durations of the youngest

reversals but cannot yet provide information of timing with reversals.

Sediment records on the other hand, generally provide a much

more continuous recording of the field. Correlation of @

O records

with the Marine Isotopic Stages can provide high-resolution age con-

trol through a reversal. Such records provide important information

regarding the age, timing, and duration of polarity transitions.

Paleomagnetic studies of polarity transitions result in a sequence of

magnetizations that were acquired as a reversal occurred. These results

are generally presented as magnetization declination, inclination, and

intensity or virtual geomagnetic pole (VGP) latitude as a function of

stratigraphic position (either as positions in a sequence of sediments

or lava flows). In order to compare records from different geographic

locations it is necessary to take into account the fact that even during

intervals of full polarity, different field directions are expected at dif-

ferent sites. For full polarity intervals, paleomagnetists do this by using

the Geocentric Axial Dipole hypothesis to calculate the equivalent

geomagnetic pole for the given magnetization directions (where the

site results would place Earth’s magnetic north pole). If the magnetiza-

tion is thought to represent a spot reading (nearly instantaneous) of the

field in time, the pole is called a VGP.

Polarity transition data are often presented as VGP positions as a func-

tion of stratigraphic position (Figure G28). This method of viewing the

data is a convenient way to show the polarity reversal by illustrating how

the apparent north magnetic pole (VGP), calculated from the observed

directions at a site, moves from one high latitude position to the other.

An additional way of presenting a transition record is to plot the

path of the VGP on a world map as it moved from one geographic pole

to the other (Figure G29). This method makes it possible to compare

records from distant sites and to test the hypothesis that the transitional

field was dipolar.

Figure G28 A record of the Lower Jaramillo polarity transition

(Gee et al., 1991) plotted as virtual geomagnetic pole position

versus stratigraphic position. The full polarity intervals exhibit

VGPs, which cluster close to the geographic poles. The dotted

intervals represent a measure of the scatter of the VGPs about

the full polarity average direction. The progression of VGP

positions from one geographic pole to the other defines the

polarity transition (shown by dashed interval). In this case, the

transitional interval is defined as the zone in which the VGPs

fall a statistically significantly distance from the geographic

poles.

GEOMAGNETIC POLARITY REVERSALS, OBSERVATIONS 325