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

Подождите немного. Документ загружается.

form. One could in principle “hide” their presence by hypothesizing

equal numbers of magnetic charges of each sign (i.e., equal numbers

of South and North poles that are spatially separated within Earth)

but in the absence of any laboratory evidence, this hypothesis has no

support.

New physical theories associated with rotation

Since Earth’s dipole is approximately aligned with the rotation axis, it

is natural to think of some (unspecified) physical process that creates a

magnetic dipole aligned with the angular momentum vector for any

rotating body. This is in the sprit of the previously mentioned gyro-

magnetic effect, except that it would be different physics (i.e., would

exist even for a nonmagnetic material) and would be important only

for very large accumulations of material. One could hypothesize, for

example, that M/J

, where M is the magnetic moment, J is the angu-

lar momentum, and n is some exponent. This proposed proportionality

has become known as the Magnetic Bode’s law (in very crude analogy

with the Titus-Bode geometric progression of planetary orbital spa-

cings), except that the label is now more often used to describe some

poorly understood scaling for the dynamo process rather than hypothe-

sized new physics. (Note that the original “Bode’s law” is not in fact a

physical law.).

The basic idea that there is some universal linear relationship

between angular momentum and magnetic moment sufficient to

explain Earth’s magnetic field was disproved in the laboratory by

Blackett during his development of highly sensitive magnetometers

in the 1950s. The strictly empirical claimed proportionality observed

when one plots M and J for planets on a log-log plot (magnetic Bodes

law) deserves to be in ill repute because it covers many orders of mag-

nitude and most of the range is explained by the fact that dipole

moments scale as planet radius cubed, while angular momenta scale

as planet radius to the fifth power. (The other parameters in this

hypothesized relationship vary by smaller amounts.) The magnetic

Bodes “law” can be used as a pedagogical example of how to mislead

oneself and others by plotting two quantities that have an underlying

common factor and thereby falsely concluding a relationship.

Summary

Table N2 summarizes nondynamo theories, their basis (or otherwise)

in fundamental physics, and their possible significance and magnitude.

With the exception of permanent magnetism, these proposals are

either real effects that are too small (usually by many orders of magni-

tude) or effects that have no established reality. The bibliography

includes reference to two papers on the planet Mercury, which illustrate

the extent to which alternatives can continue to have some credibility in

an instance where the field is much smaller than Earth and the data are

limited.

David J. Stevenson

Bibliography

Aharonson, O., Zuber, M.T., and Solomon, S.C., 2004. Crustal rema-

nence in an internally magnetized nonuniform shell: a possible

source for Mercury’s magnetic field? Earth and Planetary Science

Letters, 218: 261–268.

Barnett, S.J., 1933. Gyromagnetic effects: history, theory, and experi-

ments. Physica, 13: 241.

Elsasser, W.M., 1939. On the origin of the Earth’s magnetic field.

Physical Review, 55: 489.

Inglis, D.R., 1955. Theories of the Earth’s magnetism. Review of

Modern Physics, 27: 212.

Merrill, R.T., McElhinney, M.W., and McFadden, P.L., 1996. The

Magnetic Field of the Earth. New York: Academic Press, 531 pp.

Parker, E.N., 1979. Cosmical Magnetic Fields: Their Origin and Their

Activity. Oxford: Clarendon Press; New York: Oxford University

Press, 841 pp.

Poirier, J-P., 1991. Introduction to the Physics of the Earth’s Interior.

New York: Cambr idge University Press, p. 191.

Roberts, P.H., and Glatzmaier, G.A., 2000. Geodynamo theory and

simulations. Review of Modern Physics , 72: 1081–1123.

Russell, C.T., 1993. Magnetic fields of the terrestrial Planets. Journal

of Geophysical Research-Planets, 98: 18681–18695.

Stevenson, D.J., 1974. Planetary magnetism. Icarus, 22: 403–415.

Stevenson, D.J., 1983. Planetary magnetic fields. Reports on Progress

in Physics, 46: 555–620.

Stevenson, D.J., 1987. Mercury’s magnetic field—a thermoelectric

dynamo? Earth and Planetary Science Letters, 82:114–120.

Cross-references

Dynamos, Planetary and Satellite

Table N2 Nondynamo theories

Mechanism Physical basis Significance and possible magnitude

Primordial Finite decay time of currents Decay time is too short compared to

age of the Earth

Unexpected (noncrustal) ferromagnetism

permanent magnetism

No suitable candidate minerals exists

Gyromagnetism Dipole associated with a rotating ferromagnet Too small by at least 10

Thermoelectric effect Differences in temperature and composition Field too small and wrong geometry

Thermomagnetic effect Currents arising in presence of thermal gradients

and magnetic field

Amplification too small; no

regeneration of field

Hall effect Additional current in presence of magnetic field Small change in existing field

Rotating electric fields Current arising from rotating charged capacitor Requires impossibly large charge or

electric fields

Compression effect Gravity acts differently on electrons and massive

nuclei

Too small by 10

Magnetic monopoles Permitted in some fundamental particle physics

theories

Not observed

Magnetic “Bodes law” hypothesized

fundamental connection between

magnetic moment and angular

momentum

No observational or theoretical basis

706 NONDYNAMO THEORIES

NORMAN, ROBERT (FLOURISHED 1560–1585)

Nothing is known of the birth and death dates, parentage, or precise

marital status of Robert Norman. All that we have to go on, from infor-

mation contained in his two published works and a few other scattered

fragments, is that he served at sea for 18 or 20 years before settling in

the seafaring district of Ratcliff, on the north bank of the Thames,

London, as a maker of navigational instruments, and in particular of

marine compasses. It was to be in this latter profession that he won both

a contemporary and an enduring fame in the history of geomagnetism.

In The Newe Attractive (London, 1581), Norman published the first

serious study of the magnetic dip. For while the dip had been dis-

covered by the German Georg Hartmann around 1544, Hartmann

had only communicated his discovery in a private letter that was

not to become known to the wider world until 1831, so that Robert

Norman’s discovery, announced in print in 1581, can be rightly

credited as independent and original.

Although a working instrument maker, Norman clearly possessed

all the right instincts for an original research scientist. He made a

long-standing study of the behavior of the compass, and found that

to make a ship’s compass needle lie properly flat on its suspension

pin in London, he had to weight the south end of the needle, for it

invariably dipped down to the north. A small blob of wax was used

to act as the counter weight. Trials and experiments led him to realize

that, in London, the magnetic needle dips at an angle of 71



. This

prompted him to next investigate how the needle behaved elsewhere,

which in turn brought him to the understanding that the Earth’s mag-

netic field seemed to act in more than just the horizontal plane.

Norman realized that knowledge of the magnetic dip, when used in

conjunction with the position of the usual horizontal compass needle,

could be of great benefit to position-finding at sea. For if navigators

and researchers could build up a detailed knowledge of the position

of the horizontal and dip needles for different parts of the globe, then

a sailor could derive a better knowledge of his position. (A similar

logic was to lie behind William Gilbert’s practical uses of magnetism

and Edmond Halley’s magnetic surveys of the Atlantic, over a century

later.) Unlike Gilbert, however, Norman never developed a coherent

theory about what magnetism was or how it occurred.

Robert Norman belonged to a class of men who were coming to

play an increasingly significant role in the wider rise of science, espe-

cially in England: intelligent, ingenious craftsmen, sailors, and experi-

enced travelers, who while not “learned” in the sense of having

received a Latinate classical education, were to approach the study

of natural phenomena from a refreshing and original angle. Unfettered

as they were by classical cautions about the unreliability of sense-

knowledge, and outside the linguistic and philosophical games of the

European universities, they formed that cadre of men whose praises

were soon to be sung by Dr. William Gilbert, Sir Francis Bacon, and

the early fellows of the Royal Society. It would be wrong, however,

to see Norman, and enterprising men like him, as simple tradesmen

who were fond of writing. For, although not classically educated,

Norman seems to have had a good reading and even a translating

knowledge of Dutch, and was encouraged in his magnetical researches

by none other than William Borough, himself a student of magnetism

and comptroller of the Royal Navy, to whom The Newe Attractive

was dedicated. It is perhaps more appropriate to see Norman as a

successful entrepreneur who financed his scientific researches on the

strength of a successful business: a pattern that would be very common

in British science down to the late 19th century.

Geomagnetism, moreover, was in many ways a more challenging

field of research for such men to become occupied with, for while

mathematical-scale graduators and cross-staff makers became involved

in astronomy and oceanography, the study of magnetism was more

innately mysterious. This meant that it was much more “philosophi-

cal,” or what we would call research-physics-based, in so far that

it—like optics—involved wrestling with and discovering the mathe-

matical behavior of invisible and seemingly intangible natural forces.

Indeed, the essentially practical nature of physics research as con-

ducted by men such as Norman, he explicitly contrasted with the

approach of “learned” scholars, who preferred to remain in their stu-

dies and find answers to all things within their books. What is more,

Norman, like contemporary navigation writers such as Edward Wright,

was a pioneer of the English language as a medium of scientific and

technical discourse. His The Safegarde of Saylers, or, Great Rutter

(1584, 1590) used contemporary material, which Normon had translated

from Dutch vernacular navigation manuals, often called “Rutters” (or

Routers), and was, in fact, part of a vernacular technical and naviga-

tion book movement that included not only Dutch and English, but

also French, Spanish, and Italian materials. Indeed, all the great sea

powers of the age produced vernacular authors who wrote instruction

books and tables for their non-Latinate sea captains who now faced,

as a matter of course, long-haul oceanic voyages.

And what began as vernacular books for sailors became increasingly

the norm for learned scientists by the 17th century, and by the 1660s

the Royal Society and the French Académie regarded it as normal to

publish the researches in their mother tongue. In this way, therefore,

Norman and his magnetic researches played a role in the growth of

a new intellectual tradition. Norman’s published work also helped

to stimulate wider English researches into geomagnetism. Edmund

Gunter, Henry Gellibrand, and those early Gresham College, London,

professors who began to monitor the westward drift of the compass vari-

ation at Limehouse between 1622 and 1634 (in the wake of William

Borough’s work in the same place 50 years earlier), learned men as they

were, took at least part of their queue from Norman’s example.

Norman’s instruments were clearly prized, for Dr. Christopher

Merret, an early fellow of the Royal Society, recorded taking one of

Norman’s “Clinatories” or dip-needles for repair by Henry Bond

nearly a century after it had been made.

It is sad that the wider details of the life of a figure of such impor-

tance as Robert Norman in the early history of geomagnetism should

have been lost.

Allan Chapman

Bibliography

Bennett, J.A., 2004. Robert Norman. New DNB. Oxford: Oxford

University Press.

Chapman, A., 1998. Gresham College: Scientific instruments and the

advancement of useful knowledge in seventeenth-century England.

Bulletin of the Scientific Instrument Society, 56:6–13.

Norman, R., 1581. The Newe Attractive. London.

Taylor, E.G.R., (1954, 1968). The Mathematical Practitioners of

Tudor and Stuart England 1485–1714. London: Cambridge

University Press, p. 173 no. 29.

NORMAN, ROBERT (FLOURISHED 1560–1585) 707

OBSERVATORIES, OVERVIEW

Geomagnetic observatories carry out continuous and accurate monitor-

ing of the strength and direction of the Earth’s magnetic field over many

years, making measurements at least every minute. Observatory data

reveal how the field is changing on a wide range of timescales from sec-

onds to centuries, and this is important for understanding processes both

inside and outside the Earth. It is estimated that there are approximately

180 observatories currently operating around the world (Figure O1) (for

more information on observatories in particular countries or regions of

the world see Observatories in ... and Figure O2/Plate 3). The distribu-

tion of observatories is largely determined by the location of habitable

land and by the availability of local expertise, funds, and energy supply,

and as result, it is uneven and a little sparse in some regions. Many obser-

vatories have had to move because of encroaching urbanization. How-

ever, the continuity of their data series is generally maintained by

simultaneous observations at the old site and new site over a period of

time to allow site differences to be established.

The distribution of data in time is shown in Figure O3. Interna-

tional scientific campaigns such as the first International Polar Year

in 1882/1883, the second International Polar Year in 1932/1933, and

the International Geophysical Year in 1957/1958 encouraged the opening

of many observatories around the world. Geomagnetism is a cross-

disciplinary science, and as a result, observatories are run by a wide vari-

ety of institutes whose interests range from geology, mapping, geophysics

(including seismology and earthquake prediction), meteorology to solar-

terrestrial physics, and astronomy.

There are two main categories of instruments at an observatory. The

first category comprises variometers that make continuous measure-

ments of elements of the geomagnetic field vector but in arbitrary

units, for example millimeters of photographic paper in the case of

photographic systems or electrical voltage in the case of fluxgates

(see Observatories, instrumentation and Observatories, automation).

Both analog and digital variometers require temperature-controlled

environments and installation on extremely stable platforms (though

some modern systems are suspended and therefore compensate for

platform tilt). Even with these precautions they can still be subject to

drift. They operate with minimal manual intervention but the resulting

data are not absolute.

The second category comprises absolute instruments that can make

measurements of the magnetic field in terms of absolute physical basic

Figure O1 Locations of currently operating observatories.

Figure O2/Plate 3 Continued

OBSERVATORIES, OVERVIEW 709

units or universal physical constants. The most common types of abso-

lute instrument are the fluxgate theodolite for measuring declination

and inclination and the proton precession magnetometer for measuring

total intensity. In the former the basic unit is an angle. The fluxgate

sensor mounted on the telescope of a non-magnetic theodolite is used

to detect when it is perpendicular to the magnetic field vector. True

north is determined by reference to a fixed mark of known azimuth.

This can be determined astronomically or by using a gyro attachment.

In a proton precession magnetometer, the universal physical constant

is the gyromagnetic ratio of the proton and the basic unit is time (fre-

quency). Measurements with a fluxgate theodolite can only be made

manually while a proton magnetometer can operate automatically.

Final observatory data are produced by adjusting the variometer

data by application of a baseline, so that they closely fit the absolute

data. In this process any long-term drifts or steps in the variometer data

and differences between the sites of the variometers and absolute

Figure O2/Plate 3 The Danish Meteorological Institute suspended triaxial fluxgate magnetometer being adjusted at Alibag

observatory in India (a). This variometer can measure the magnetic field variations along the axes of three orthogonal sensors every

second. Another instrument also run in continuous mode at observatories is a proton precession magnetometer. It is an absolute

instrument and measures the strength of the magnetic field. In order to convert the variometer data into absolute data, manual

observations are necessary. For this a fluxgate theodolite, or DI-flux, is widely used. (b) Shows the fluxgate theodolite at Kakadu

observatory in Australia. The fluxgate sensor is fixed to the telescope of the non-magnetic theodolite and is used to indicate when its axis

is orthogonal to the magnetic field, and therefore the orientation of the field (see Observatories, instrumentation and Observatories,

overview). (c) Fu

rstenfeldbruck observatory, Germany; (d) Resolute observatory, Canada; (e) Qeqertarsuaq (Godhavn) observatory,

Greenland; (f) Tamanrasset observatory, Algeria; (g) Hartland observatory, UK; (h) Dumont D’Urville observatory, Antarctica; (i) Tihany

observatory, Hungary. These buildings, some in the vernacular style, are made of non-magnetic material and can be maintained at stable

temperatures if housing variometers.

Figure O3 Number of observatory annual means per year in the Northern (grey) and Southern (black) Hemispheres.

710 OBSERVATORIES, OVERVIEW

instruments are effectively removed. Various other instrument effects

may also be included in the baselines. These are scaling factors, offsets,

sensor alignments, temperature responses, and timing errors. Many of

these effects are insignificant with modern digital variometers. If there

is a backup variometer, ideally housed in a different building on

the observatory site, filling of gaps resulting from human-made noise

and data loss may also be possible. Once a continuous clean time series

of data reduced to the observatory reference location is obtained, final

observatory products are produced and disseminated. These include

1-min means, hourly means, annual means, and K-indices.

There have been many efforts to standardize the operation of mag-

netic observatories and to expand the network. Current international

efforts include the IAGA Observatory Workshops and INTERMAG-

NET but there are also many efforts made by individual institutes

(see Observatories in ...). The International Association of Geomag-

netism and Aeronomy (see IAGA, International Association of Geo-

magnetism and Aeronomy) runs workshops every 2 years where

information is exchanged on observatory practice and instrumentation.

IAGA has also published a guide to observatory practice (Jankowski

and Sucksdorff, 1996). Since the late 1980s INTERMAGNET (Inter-

national Real-time Magnetic observatory Network) has contributed to

the establishment of global standards in the measurement, recording

and dissemination of digital geomagnetic data (see Observatories,

INTERMAGNET ). There are now about 100 observatories operating

to INTERMAGNET standards.

The satellite magnetic survey missions of the International Decade

of Geopotential Research, which started with the launch of the Ørsted

satellite in 1999 (see Ørsted and CHAMP), along with increased

appreciation of the effects of space weather on technological systems

during magnetic storms (see Geomagnetic hazards) has fueled recent

interest in geomagnetic observatory data. The many programs to

replace analog systems with digital systems will also increase the uti-

lity of observatory data. The future for geomagnetic observatories is

hopefully a bright one.

Susan Macmillan

Bibliography

Jankowski, J., and Sucksdorff, C., 1996. IAGA Guide for Magnetic

Measurements and Observatory Practice. International Association

of Geomagnetism and Aeronomy.

Cross-references

CHAMP

Geomagnetic Hazards

IAGA, International Association of Geomagnetism and Aeronomy

Observatories, Automation

Observatories in ...

Observatories, Instrumentation

Observatories, INTERMAGENT

Observatories Program in ...

Ørsted

OBSERVATORIES, INSTRUMENTATION

Introduction

The main task of a geomagnetic observatory is to observe the natural

geomagnetic field for an extended period of time (at least 1 year) by

performing continuous measurements of the three-dimensional geo-

magnetic vector. Observatory instrumentation is responsible for this

measurement task. As we are dealing with a very wide-frequency spec-

trum of signals, whose characteristic times extend from centuries to

subsecond intervals (see Geomagnetic spectrum, temporal), issues of

long-term instrument reliability, standardization of the instrumenta-

tion, continuity in data formats, as well as fast sampling capability

are to be considered carefully. As only the natural geomagnetic field

is of interest here, observatory instrumentation should be completely

nonmagnetic: it should not modify the direction or amplitude of the

geomagnetic vector in a measurable way. If, however, the sensors are

magnetic, or produce magnetic fields for the sake of the measurement

procedure, it must be ensured that those effects are eliminated by

the measurement protocol and that no perturbation on neighboring

instruments arises.

For the instrumentation to perform successfully, the observatory

must answer to strict magnetic conditions: low magnetic gradient

and identity of field changes over the area. Moreover, one is led to

consider the whole observatory site as part of the observatory instru-

mentation; hence the quality of measurement depends on the magnetic

hygiene of the surroundings and the buildings, on low magnetic noise

conditions, as well as on the stability of the pillars bearing the field

orientation sensors. Detailed information about geomagnetic instru-

mentation in observatories can be found in Jankowski and Sucksdorff,

1996.

Short history

The first magnetic observatory instrument—the magnetic compass—

was discovered in China already before the year

A.D. 1000. Early

observatories—starting in the 16th century—used an oversize com-

pass needle so as to ensure sufficient resolution on the declination

angle. The compass was later complemented by the dip circle for mea-

suring the magnetic inclination. Still later, projected field components

would be measured by using setups with current-carrying coils and

calibrated magnets. A (virtual) visit to the Museum of the History

of Science in Florence will allow you to see those early instruments.

Continuous observations were performed by an observer making a

reading every 2 h or so and writing the values down in a notebook.

The photographic paper recorder appeared in the 19th century, allow-

ing uninterrupted graphical registration of the field. The invention of

the fluxgate in the early 1900s permitted the development of the triax-

ial electronic variometer and the DI-flux, which are in use today in

most magnetic observatories. The introduction of the proton preces-

sion magnetometer (Packard and Varian, 1954) allowed an important

operational simplification and improvement in the measurement accu-

racy of the magnetic induction (see Instrumentation, history of ).

Geomagnetic metrology

To completely describe the geomagnetic vector, each measurement

must sample three independent vector components. The three most

popular coordinate systems for describing the geomagnetic vector are:

Cartesian: X (North component), Y (East component), Z (vertical

component, positive downward).

Cylindrical: D (magnetic declination angle, positive East), H (hori-

zontal component).

CRLF: Z Spherical: D, I (magnetic inclination angle, positive down-

ward), F (magnetic field induction intensity).

Component magnetometers will measure projections of the vector

with a large constant part, like X and Z, while orientation magnet-

ometers will be sensitive to the orientation of the field with respect

to the sensor and measure small projections like D or Y. Component

and orientation magnetometers are not yet capable of continuously

measuring all the field components with the required absolute accu-

racy. Therefore, an observatory measurement setup uses:

 a variometer to measure the variation of the field components

about baseline values in a continuous and unattended way at the

required sampling rate, say 1 Hz and

 absolute measurements performed regularly (say 2 per week) by an

observer with the adequate instrumentation (DI-flux, proton mag-

netometer; see Figure O4) to establish the values of the baselines.

OBSERVATORIES, INSTRUMENTATION 711

This makes the observatory measurement essentially a two-step pro-

cess, where the second part is performed manually. Postprocessing

merges the two data sets to produce a final data record having the

accuracy of the absolute instruments at the variometer’s sampling rate.

The expression “absolute measurement” means here that the observa-

tions must be fully traceable to metrological standards for the magnetic

induction and that the orientation of the geomagnetic vector is mea-

sured with respect to the local vertical and to geographic north.

It is expected that in the future fully automatic observatories will be

available (see Observatories, automation) where the absolute measure-

ment part will be also unattended. Observers are striving to make abso-

lute measurements with an accuracy of 0.1 nT for inductions and 1 s of

arc for orientation. The INTERMAGNET consortium (see Observa-

tories, INTERMAGNET) has laid down observatory instrumentation

specifications for its members. This conveniently provides guidelines

for observatories wishing to upgrade their instrumentation.

In less-modern observatories, variometers consist of suspended

magnets and magnetic balances (Wienert, 1970). The continuous

record of the field’s components is recorded as curves on a piece of

photographic paper, the magnetogram. Elongated axial magnets, sus-

pended so as to rotate freely around the vertical, will orient their mag-

netic axis along the geomagnetic meridian, and will thus indicate its

direction, providing a declination variometer. If an auxiliary stationary

magnet creates a bias field so as to position the suspended magnet with

its magnetic axis east-west, the suspended magnet’s motion will follow

the changes in the horizontal component, allowing it to be recorded.

Magnetic balances compare the torque exerted by the vertical compo-

nent on a magnet with the one exerted by gravity on a constant mass

using a knife-edge balance. In all cases, the recording of the magnet’s

motion is ensured by a light beam falling on a mirror at the magnet’s

extremity. The horizontal beam, deflected by this mirror, will impress a

time-dependent curve on a vertically streaming photographic paper.

The streaming speed is usually about 20 mm h

1

, but for rapid run

magnetographs it may go up to 200 mm h

1

In modern observatories, the most popular magnetometers are based

on the proton precession effect and the fluxgate sensor. In former

Soviet bloc countries, the Bobrov magnetometer, which consists of a

magnet suspended by a taut quartz fiber, is successfully used in its

servo version, where the magnet’s orientation is kept fixed by the

action of a coil. The coil current is easily fed to a digital data acquisi-

tion system to produce computer readable files.

The proton precession magnetometer

It is used in observatories for measuring the magnetic induction B in

the range of 10–100 mT with accuracy down to 0.1 nT. Here the free

precession frequency f of protons is the observable and the formula

2pf ¼ g

B (Eq. 1)

allows the measurement of the magnetic induction by way of the mea-

surement of a frequency. The standard for the magnetic induction is

thus conveniently converted to a frequency standard, which is widely

available. The quantity g

is the proton gyromagnetic ratio at low field

for a spherical H

O sample at 25



C. The recommended value given by

CODATA and adopted by IAGA in 1992 is (Cohen and Taylor, 1987)

¼ 2:67515255T

1

(Eq. 2)

The sensor consists of a bottle containing the proton-rich fluid and is

surrounded by a coil, serving the dual purpose of applying periodically

a polarizing field to the liquid and picking up the signal from the pre-

cessing protons after cutting off the polarizing field. An electronic con-

sole will amplify the precession signal and perform a frequency

measurement of it with the required accuracy. This measurement is

then scaled using g

to give the field induction intensity in teslas.

This quite remarkable instrument is one of the few sensors able to

measure directly the modulus of a vector without using computation

from components or performing a leveling or tedious orientation pro-

cedure. This is of course due to the fact that the free protons have only

the ambient magnetic field to orient themselves.

Modern proton magnetometers can be quite compact and have high

resolution and sampling rate (0.01 nT and 3 Hz are typical). The Over-

hauser types (Sapunov et al., 2001) in particular are responsible for

this progress and have the additional benefit of low power operation.

They are therefore a first choice in modern observatories instrumenta-

tion, providing automatic and continuous observation of the geomag-

netic induction field intensity.

The optically pumped magnetometer

Optically pumped magnetometers (OPM) are based on the splitting of

the energy levels of some alkali metal and helium atoms in a magnetic

field. The optical pumping scheme allows measuring this energy split-

ting with very high resolution (Alexandrov and Bonch-Bruevich,

1992).

OPMs have been used at magnetic observatories, both for total field

modulus measurement and, when complemented by a set of bias coils,

for observing the geomagnetic field direction changes. One of the first

digital observatories was based on a rubidium OPM (Alldredge and

Saldukas, 1964).

Like proton magnetometers, the OPMs are scalar magnetometers

but, to the contrary of the proton magnetometer, deliver a continuous

stream of data, in the form of a frequency modulated by the value of

the field. This frequency, with name of Zeeman, is linked to the mea-

sured magnetic induction by the Breit-Rabi polynomial formula. In a

few cases, when the Zeeman peaks corresponding to the different

energy-splits in the atoms are resolved, the polynomial coefficients

can be calculated from fundamental physical constants, and the OPM

has then absolute accuracy. This is the case of the OPMs based on

potassium,

He, and

He vapor (Alexandrov and Bonch-Bruevich,

1992; Gilles et al., 2001).

OPMs have a high sensitivity with some approaching the 0.1 pT

1/2

noise level. This performance makes them attractive for pro-

spection work, aeromagnetic survey, space-based observations, and

field stabilizers. However their high initial and maintenance cost due

to the short lifetime of the gas-discharge pumping lamp has limited

their use in the observatory. It looks like this deficiency will be over-

come in the future in view of quick progress in laser-pumping lamps

(Gilles et al., 2001).

Figure O4 Instrumentation as used in most magnetic

observatories of the world. The left panel represents a triaxial

fluxgate variometer sensor. Right panel: the DI-flux magnetometer

mounted on top of the pillar consists of a nonmagnetic theodolite

with a fluxgate sensor on its telescope. In the foreground, an

Overhauser proton magnetometer consisting of the cylindrical

sensor and the electronic console.

712 OBSERVATORIES, INSTRUMENTATION

Induction coil magnetometer

The induction coil (or search-coil) magnetometer’s (ICM) operation

principle is based on the Maxwell equation which can be written in

integral form as

e ¼

;

where e is the electromotive force (emf ), F is the magnetic flux, and

t is the time. In its simplest form an ICM contains a coil with many

turns of copper wire connected to the input of a voltage amplifier.

As it is seen from the equation, an ICM can be used only for the mea-

surement of a time-varying magnetic flux. Its component collinear to the

coil’s axis intercepts the coil loops and generates an emf at the coil’s

terminals, which is further amplified to an easily measurable level.

To increase the ICM sensitivity, a high permeability ferromagnetic

material is used as a core inside the multiturn winding. Due to their

relative simplicity, ICMs are widely used for many applications,

mainly in geophysics. In geomagnetic observatories they are used for

the study of Earth’s magnetic field pulsations.

Their operational frequency band covers from about 10

–4

till 10

þ7

Hz and the measurement dynamic range covers the range from frac-

tions of femtotesla till tens of tesla. In spite of an apparent simplicity,

the creation of a high-class ICM needs complicated calculations to

establish an optimal matching of the sensor coil with the amplifier

(Korepanov et al., 2001).

The fluxgate magnetometer

This instrument is used for measuring the component of the magnetic

field vector along its sensor’s axis. The fluxgate principle uses the

nonlinear field/induction relationship of an easily saturable ferromag-

netic core. The core, usually a rod or a ring, is subjected to both the

DC field to be measured, and an auxiliary AC field produced by a coil

and an electronic oscillator. This offset sinusoidal excitation will create

a distorted AC signal, in a pick-up coil surrounding the core. The

detection of its even harmonics provides a DC signal proportional to

the field to be measured.

Fluxgate sensors work best in small axial fields, therefore they are

often operated within a third current-carrying coil, which cancels off

the main part of the DC field. This main part may be calculated by a

servo scheme, or be constant. Otherwise, the fluxgate sensor is oper-

ated with the field essentially normal to the sensor, resulting in a small

field component along the core axis. The fluxgate will then be sensi-

tive to its orientation to the field.

Observatory fluxgate variometers will include three fluxgate sensors

arranged orthogonally on a stable support made of marble or quartz.

This trihedron is then oriented by the variometer frame according to

the three components set one wishes to observe (see Figure O4).

Absolute meas urements of declination and inclination

Measuring the orientation of the geomagnetic vector with respect to a

coordinate frame pointing to the geographic north and to the local ver-

tical represents a metrological challenge, especially if one wishes to

attain 1

direction accuracy on this fluctuating vector. In fact, several

measurements are taken so as to correct for systematic instrumental

defects. Therefore the variometer of the observatory is used for taking

into account any field changes during the measurement procedure.

The state of the instrumentation art is now provided by a device called

“DI-flux” or “DIM,” which is assembled from a nonmagnetic theodo-

lite and a fluxgate sensor mounted on the telescope. The magnetic axis

of the fluxgate should be nearly parallel to the optical axis of the tele-

scope (see Figure O4). With a DI-flux so configured, a measurement

protocol suggested by Lauridsen (1985) allows the error-free determi-

nation of D and I.

Jean L. Rasson

Bibliography

Alexandrov, E.B., and Bonch-Bruevich, V.A., 1992. Optically pumped

magnetometers after three decades. Optical Engineering, 31:

711–717.

Alldredge, L.R., and Saldukas, I., 1964. An automatic standard

magnetic observatory. Journal of Geophysical Research, 69:

1963–1970.

Cohen, E.R., and Taylor, B.N., 1987. The 1986. CODATA recom-

mended values of the fundamental physical constants. Journal of

Research of the National Bureau of Standards (U.S.) , 92:85–95.

Gilles, H., Hamel, J., and Chéron, B., 2001. Laser pumped

He mag-

netometer. Review of Scientific Instruments, 72: 2253–2260.

Jankowski, J., and Sucksdorff, C., 1996. Guide for Magnetic Measure-

ments and Observatory Practice. Boulder: International Associa-

tion of Geomagnetism and Aeronomy.

Korepanov, V., Berkman, R., Rakhlin, L., Klymovych, Ye., Pristai, A.,

Marussenkov, A., and Afanassenko, M., 2001. Advanced field

magnetometers comparative study. Measurement, 29: 137–146.

Lauridsen, K.E., 1985. Experiences with the DI-fluxgate magnet-

ometer inclusive theory of the instrument and comparison with

other methods. Danish Meteorological Institute Geophysical

Papers, R-71.

Packard, M., and Varian, R., 1954. Free nuclear induction in the

Earth’s magnetic field (abstract only). Physical Review, 93: 941.

Sapunov, V., Denisov, A., Denisova, O., and Saveliev, D., 2001. Pro-

ton and Overhauser magnetometers metrology. Contributions to

Geophysics & Geodesy, 31:119–124.

Wienert, K.A., 1970. Notes on Geomagnetic Observatory and Survey

Practice. Brussels: UNESCO.

Cross-references

Compass

Gauss’ Determination of Absolute Intensity

Geomagnetic Spectrum, Temporal

Instrumentation, History of

Magnetometers, Laboratory

Observatories, Automation

Observatories, INTERMAGNET

OBSERVATORIES, AUTOMATION

In 1847 magnetic recording by photography was introduced at the

Royal Observatory at Greenwich. This early example of observatory

automation was prompted by the difficulty and expense of maintaining

a schedule of hourly observations coupled with the scientific desire to

see a fuller spectrum of magnetic variations. Economics is a major fac-

tor behind the present-day move to automate as many observatory

operations as possible. If Canada, for example, were to place a full-

time observer at each of its 13 observatories, the cost of operating

the network would roughly double. Fully automated observatories

can be deployed in remote locations where staff is unavailable, includ-

ing the ocean bottom. Automation is also a natural consequence of the

necessity for observatories to meet the present-day scientific require-

ments for digital data. Digital magnetometer systems naturally lend

themselves to automation.

All aspects of observatory operations are subject to full or partial

automation: (1) data collection, (2) data telemetry, (3) data processing,

(4) data dissemination, (5) error detection, and (6) absolute observa-

tions. The natural end point of the automation process is an automatic

magnetic observatory (AMO). Not all aspects of observatory opera-

tions need to be completely automated for the observatory to be con-

sidered an AMO. However, (1), (2), and (6) must be fully automated

and (3), (4), and (5) must be at least partially automated. Although

OBSERVATORIES, AUTOMATION 713

several systems previously or currently in use have been called

“automatic observatories” none truly deserves the designation.

Observatories may be classified as fully staffed if there is an operator

on duty everyday, unstaffed if the observatory is visited less than once

per month, or partially staffed if visits to the observatory fall between

these two extremes. A fully staffed observatory may be highly auto-

mated; an unstaffed observatory must be highly automated. Regardless

of the level of staffing or the degree of automation, the quality of the

data obtained from the observatory should meet accepted international

standards such as those promulgated by INTERMAGNET. In particular,

the final data disseminated by the observatory should have an absolute

accuracy of better than 5 nT, so that data may be used to study secular

variation and other long-period changes in the magnetic field. The

requirement for absolute accuracy, as opposed to relative accuracy, dis-

tinguishes magnetic observatories from other magnetometer installa-

tions. To define absolute accuracy consider the difference between the

true value of the magnetic field, B, and the reported value, Br;given

by d ¼ E þ s where E denotes a slowly varying or systematic error

and s denotes a random error. To achieve high absolute accuracy, E is

treated as a parameter that can be determined through a series of calibra-

tion, or “absolute”, observations that are used to correct the output

values of the magnetometer. If data are only used to study short-period

variations, relative accuracy, rather than absolute accuracy, is important.

We assume that E remains constant over the time interval t

–t

,sothat

the difference Br(t

)–Br(t

) will accurately represent B(t

)–B(t

), pro-

vided s is small, even though E, and therefore B(t

)andB(t

) are not

accurately known.

Alldredge (1962) designed a highly automated system, ASMO

(Automatic Standard Magnetic Observatory) which was intended to

produce outputs similar to those obtained by scaling standard magne-

tograms. However, since 1-min values were recorded to tape, it was

obvious that many additional types of analyses could be carried out.

ASMO consisted of a rubidium magnetometer located at the center of

a two-axis Helmholtz coil system. This was oriented such that the total

vector field would be biased to produce sequential changes in the angle

of declination and inclination. When installed at Castle Rock, California,

the system featured data transmission via telephone line to a receiving

and recording center 160 km away (Blesch, 1965). Thus, it fulfilled

several, but not all, of the requirements of an AMO.

The fluxgate magnetometer rapidly superceded the vector proton

magnetometer as the primary instrument in automated systems. For

example, AMOS, deployed in Canada in 1969 (Delaurier et al.,

1974), consisted of a triaxial fluxgate sensor mounted inside a square

Helmholtz coil system so that the sensor essentially operated in zero

field. D, H, Z (and F ) were sampled every minute and recorded on

tape. By 1974, it was agreed that the system was at least as reliable

as the photographic recorder it was intended to replace and could in

fact record large, dynamic magnetic storms, which the photographic

system was incapable of recording. AMOS featured an innovative

telephone verification system (TVS) that permitted remote diagnosis

of malfunctions. Given the lack of stability of fluxgate magnetome-

ters of that era, the deployment of AMOS did not eliminate or reduce

the need for absolute observations. There was no longer a need for

full-time personnel at the observatories, but part-time contractors were

still needed to perform absolute observations and other minor duties.

Thus, of the six areas of automation, AMOS addressed only the areas

(1) and (5).

Other early systems that incorporated fluxgate magnetometers

include those installed at Erdmagnetisches Observatorium Wingst in

1980 (Schultz, 1983), at British observatories in 1979 (Forbes and

Riddick, 1984) and at French high (southern) latitude observatories

in 1972 and at Chambon-la-Forêt (Bitterly et al., 1986).

By the mid-1980s advances in magnetometry, computers, and teleme-

try increased the possibilities for automation. Although developments

took place in many countries, the UK and United States (quickly joined

by France and Canada) spearheaded a movement to establish modern

standards for digital magnetometers and data. The initiative, which

evolved into INTERMAGNET, also promoted rapid telemetry and auto-

matic dissemination of data. Any observatory that adopts INTERMAG-

NET standards and specifications will possess a system capable of

providing digital data that can be collected, processed, and distributed

automatically. Such an observatory will typically possess a late genera-

tion fluxgate magnetometer system controlled by and recording data to a

computer or other microprocessor-controlled data collection platform

(DCP). Data will be filtered in the DCP to form 1-min values, which

will be transmitted by satellite or phone line to the host institute and/

or an INTERMAGNET geomagnetic information node (GIN), where,

after some basic error checking, they can be obtained automatically

by users. Thus an INTERMAGNET observatory is likely to achieve

complete automation of (1), almost complete automation of (2), and at

least some automation of (3) and (4).

One of the first such systems was ARGOS (Automatic Remote

Geomagnetic Observatory System), which became operational at

the three UK observatories in 1987. ARGOS consisted of a three-

component fluxgate and a proton precession magnetometer controlled

by a computer, which also acted as the data logger. Data were trans-

mitted to a central office by phone line. The US system was similar

in concept; however, it used a ring core fluxgate and the (GOES) for

transmission of data. The Canadian system, Canadian magnetic obser-

vatory system (CANMOS), used the same fluxgate magnetometer

as the US system and also used the GOES for data transmission. It

also included an improved remote diagnostic system. The French

system included an improved magnetometer operating in feedback

coils. Similar automated magnetometer systems are now (2005) in

operation at approximately 100 observatories.

Subsequent improvements to magnetometers and data collection sys-

tems have been concerned with magnetometer stability and sampling

rates. Efforts have been made to make the output of the fluxgate mag-

netometer as stable as possible, thus reducing the need for independent

absolute observations. Tilt-reducing sensor suspensions are capable of

eliminating the effects of pier movement. Temperature coefficients of

fluxgate magnetometers typically range from 0.1 to 1 nT K

1

(Rasson,

2001) leading to errors of up to 10 nT. Temperature coefficients can be

determined for some systems but for others it is simpler to keep the sys-

tem at a constant temperature. Most observatories recognize the need of

the scientific community for data at a faster sampling rate than 1-min;

many are now providing 1-s data.

Many observatories currently rely on GOES, Meteosat, or GMS for

telemetering 1-min data to the host institute. However, this method of

telemetry is incapable of coping with the increasing need for near-realtime

1-Hz data. The Internet, which is increasingly accessible at even remote

locations, is becoming the method of choice for many institutes, which

operate networks with satellite and phone systems serving as backup.

It is unlikely that data processing will ever be carried out without at

least some human intervention. There will always be instances when

judgment is necessary to decide on the validity of a datum or series of

data. A certain amount of data processing may be carried out in the

DCP; the rest is normally carried out after the data are received at the

host institute. Tasks that have been automated include removal of large

spikes, plotting of magnetograms, production of indices, computation of

hourly means, and computation and addition of preliminary baselines.

More difficult and perhaps impossible to completely automate are such

tasks as detection and removal of small spikes, especially at auroral lati-

tudes; handling of offsets; and calculation of final baselines, including

the detection and removal of spurious absolute observations.

Sodankylä and Tromsö were the first observatories to use the Inter-

net to display and download data (Linthe, 2001). Today many observa-

tories have taken advantage of the World Wide Web to provide access

to their data and products such as magnetogram plots, minute data

(and faster sampled data as well), hourly, monthly, and annual mean

values, indices, and observatory and network descriptions.

Automated error detection and diagnosis is extremely important

when an institute runs a network of unstaffed or partially staffed obser-

vatories. Frequently the part-time operator at the site can be instructed

714 OBSERVATORIES, AUTOMATION

on how to fix a problem once the nature of the problem has been

established. In addition, some problems can be corrected remotely

without on site human intervention. Tasks that can be performed remo-

tely include rebooting the computer, adjusting the time, changing

temperature of an instrument enclosure, adjusting pier differences,

removing spikes or periods of bad data, and erasing old files. Para-

meters that can be monitored include temperature, humidity, pier tilt,

and DF, the difference between F measured with a proton magnet-

ometer and F calculated from the values of a vector magnetometer.

To be considered a true AMO an observatory must either automate

its method of performing absolute observations or, alternatively, con-

struct as system that is so stable that absolute observations need be car-

ried out infrequently. With one notable exception, the vector proton

magnetometer (VPM) has been the instrument of choice for either of

these approaches. One of the most successful of these has been Kasm-

mer (Kakioka Automatic Standard Magnetometer), first installed at

Kakioka Observatory in 1973. Both the British and US systems have

employed VPMs, known as DIDD (delta I-delta D) as a means of pro-

viding baseline control for the fluxgate magnetometers. Recently the

US Geological Survey and the Eötvös Loránd Geophysical Institute

of Hungary have collaborated to produce a highly stable DIDD featur-

ing a compact spherical coil design. Evaluations of the system have

proved encouraging (Pankratz, 1998; Csontos et al., 2001; Schott

et al., 2001). Nevertheless, the DIDD is still susceptible to pier tilt

and other forms of misalignment and can therefore be considered

only quasi-absolute. Independent absolute observations are still

required to meet the standards mentioned earlier. However, in a region

otherwise devoid of observator ies, one equipped with a good quasi-

absolute DIDD would certainly be better than no observatory at all.

Meanwhile, efforts continue to improve VPMs. In France, a

He pumped

magnetometer is being developed which will eliminate the need for a

coil system (Gravrand et al., 2001); consideration is also being consid-

ered to developing a tilt-compensating suspension for the Hungarian

or US system.

A totally different approach to automating absolute observations

has been taken by Rasson (1996): the automation of a declination-

inclination magnetometer (DIM). The DIM consists of a single-axis

fluxgate sensor mounted on a nonmagnetic theodolite; by rotating and

tumbling the telescope in a prescribed manner the angle of declination

and inclination can be obtained. Rasson has been able to automate this

process through the use of piezoelectric motors, digital readout of the

circles, accurate recording of tilt and the use of an autocollimator for

sighting the reference mark. Development is also underway of an alter-

native version using a north-seeking gyroscope to eliminate the need for

a reference mark; this would potentially make it possible to install the

device on an ocean bottom observatory magnetometer, thus fulfilling a

long-standing dream of the observatory community.

Lawrence R. Newitt

Bibliography

Alldredge, L.R., 1962. A Proposed Automatic Standard Magnetic

Observatory. U.S. Department of Commerce, Coast and Geodetic

Survey, Washington.

Bitterly, J., Cantin, J.M., Burdin, J., Schlich, R., Folques, J., and

Gilbert, D., 1986. Digital recording of variations in the Earth’s

magnetic field in French observatories: Description of equipment

and results for the period 1972–86. In Proceedings of the Interna-

tional Workshop on Magnetic Observatory Instruments, Geological

Survey of Canada Paper 88-17: 75 –81.

Blesch, J., 1965. An automatic magnetic observatory. Geophysical

Technical Memorandum No. 21, Varian Associates.

Csontos, A., Hegymegi, L., Heilig, B., and Kõmendi, A., 2001. The

test results of the Delta AI@ Delta AD@ (DIDD) measurement

system at the Tihany Geomagnetic Observatory of Elgi. Contribu-

tions to Geophysics and Geodesy, 31:83–89.

Delaurier, J.M., Loomer, E.I., Jansen van Beek, J., and Nandi, A.,

1974. Editing and evaluating recorded geomagnetic components at

Canadian observatories. Publication of the Earth Physics Branch,

44(9):235–242.

Forbes, A.J., and Riddick, J.C., 1984. The digital recording system

operated at the U.K. magnetic observatories. Geophysical Surveys,

6: 393–405.

Gravrand, O., Khokhlov, A., Le Mouël, J.-L., and Léger, J.M., 2001.

On the calibration of a vectorial,

He pumped magnetometer. Earth

Planets Space, 53: 949–958.

Linthe, H.-J., 2001. How geomagnetic observatories present collected

data. Contributions to Geophysics and Geodesy, 31: 151–158.

Pankratz, L.W., Test Results on a New Hungarian/US Delta I–Delta D

(DIDD) Quasi-absolute Spherical Coil System, 1998. VIIth IAGA

Workshop on Magnetic Observatory Instruments, Data Acquisition

and Processing, Scientific Technical Report STR98/21, 147–149

GeoForschunsZentrum, Potsdam.

Rasson, J.L., 1996. Report on the progress in the design of an auto-

mated diflux. In Proceedings of the VIth Workshop on Geomag-

netic Observatory Instruments Data Acquisition and Processing,

Publication scientifique et technique No. 003, Institut Royal

Meteorologique-De Belgique, pp. 190–194.

Rasson, J.L., 2001. The status of the world-wide network of magnetic

observatories, their location and instrumentation. Contributions to

Geophysics and Geodesy, 31: 427–439.

Schott, J.J., Boulard, V., Pérès, A., Cantin, J.M., and Bitterly, J., 2001.

Magnetic component measurements with the DIDD. Contributions

to Geophysics and Geodesy, 31:35–42.

Schultz, G., 1983. Experience with a digitally recording magnetometer

system at Wingst geomagnetic observatory. Deutsche Hydrogra-

phische Zeitschrift, 36: 173–190.

Cross-references

Observatories, INTERMAGNET

Observatories, Overview

OBSERVATORIES, INTERMAGNET

INTERMAGNET (International Real-time Magnetic Observatory

Network) was created in order to establish a worldwide network of

cooperating digital magnetic observatories. Those observatories agree

to adopt modern standard specifications for measuring and recording

equipment and to exchange data in close to real time. Moreover,

INTERMAGNET extends technical support for maintaining and

upgrading existing magnetic observatories as well as for establishing

new ones.

INTERMAGNET defines standards for the measurement and

recording of the geomagnetic field, considering the state of the art

(see Observatories, instrumentation). INTERMAGNET is constituted

from existing groups whose task is geomagnetic observatory measure-

ment. The abbreviation IMO is used to indicate an INTERMAGNET

Magnetic Observatory (Green et al., 1998).

Short history

The idea of a near real-time observatory network arose independently

from discussions inside the Geological Survey of Canada and between

US Geological Survey (USGS) and British Geological Survey in 1986

during an IAGA workshop in Ottawa (see IAGA, International Asso-

ciation of Geomagnetism and Aeronomy). The idea was expanded to

a global concept at the International Union of Geophysics and Geodesy

(IUGG) assembly in Vancouver in 1987 as Institut de Physique du Globe

de Paris joined. The network was based on magnetic observatories com-

municating as meteorological data collecting platforms and transmitting

OBSERVATORIES, INTERMAGNET 715