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

Подождите немного. Документ загружается.

Primary and secondary remanence

When igneous rocks and baked archaeological materials cool on forma-

tion, they acquire a primary thermoremanent magnetization (TRM)

(q.v.) in the ambient geomagnetic field as the temperature falls below the

Curie temperature of the magnetic minerals. A later thermal event below

the Curie point will unblock part of the TRM and replace it with a later

partial thermal remanent magnetization (pTRM) secondary remanence.

If magnetic minerals grow in the presence of a magnetic field, a chemical

remanent magnetization (CRM) (q.v.) is formed when a critical blocking

volume is attained. The remanence in sedimentary rocks is often a

CRM. CRMs can also be produced in igneous and metamorphic rocks.

In sediments, the primary remanence is usually a depositional remanent

magnetization (DRM) (q.v.) acquired when ferromagnetic particles are

aligned in the geomagnetic field during deposition. In sedimentary

rocks, dewatering during the prolonged process of lithification will

usually have enhanced the alignment of these grains with the geomag-

netic field to produce a postdepositional detrital magnetization (PDRM).

Samples may acquire a magnetization at ambient temperatures over

prolonged lengths of time in a weak magnetic field such as the Earth’s.

This is called viscous remanent magnetization (VRM) (q.v. ) and its

importance is a function of time spent in the field, the strength of the

field, and the grain size and composition of the material. VRM is the

most common form of secondary remanence and is present, to varying

degrees, in all palaeomagnetic samples. Lightning reaching the ground is

capable of inducing magnetization in rocks in a very strong, essentially

instantaneous magnetic field. This is called an isothermal remanent magne-

tization (IRM) (q.v.) and can often completely obscure the primary rema-

nence. Meteorites are extraterrestrial samples that can acquire a secondary

shock remanent magnetization (SRM), both in the event that excavated

the material from the parent body and during impact at arrival on Earth.

Provided the rock has isotropic properties, each of these types of

magnetization is in the direction of the magnetic field at the time of

formation. The intensity of magnetization is dependent on many things

(for example, the number, composition, and grain size of magnetic

minerals) and also depends on the strength of the magnetic field.

It is only possible to determine the absolute intensity of the field from

samples containing a TRM (see Palaeointensity).

Isolation of NRM components

To isolate the different components of NRM, partial demagnetization

techniques are used, as the different components will normally have

different stabilities to the demagnetization procedures. The techniques

used to resolve the components are thermal and alternating field (AF)

demagnetization. Components acquired by differing mechanisms will

usually have contrasting blocking temperature and coercivity spectra.

During thermal demagnetization, samples are heated to successively

higher temperatures in a magnetic field-free space so that magnetiza-

tion below the treatment temperature is unblocked and randomized.

As the temperature increases, more and more of the remanence will be

removed until the Curie temperature is reached and the sample is com-

pletely demagnetized. With AF demagnetization, the peak AF field is

increased until magnetic minerals with the highest coercivity are demag-

netized (or maximum laboratory peak AF field is reached). Figure M115

shows an example of thermal and AF demagnetization. The characteris-

tic remanent magnetization (ChRM) is the earliest acquired component

of magnetization that can be isolated. So for both examples in Figure

M115 this is the primary TRM that the basalt acquired on cooling.

Mimi J. Hill

Bibliography

Butler, R.F., 1992. Paleomagnetism: Magnetic Domains to Geologic

Terranes. Boston, MA: Blackwell Science.

Tarling, D.H., 1983. Palaeomagnetism Principles and Applications in

Geology, Geophysics and Archaeology. London: Chapman & Hall.

Cross-references

Magnetization, Chemical Remanent (CRM)

Magnetization, Depositional Remanent (DRM)

Magnetization, Isothermal Remanent (IRM)

Magnetization, Thermoremanent (TRM)

Magnetization, Viscous Remanent (VRM)

Paleointensity

MAGNETIZATION, OCEANIC CRUST

In the late 1950s, the first magnetometers adapted to sea surface mea-

surements became available for the scientific community, leading to

the discovery of magnetic lineations on the oceanic crust. Since this

discovery, first described by Fred Vine and Drummond Matthews in

1963 (the work of Morley and La Rochelle at the same time should

also be mentioned), the nature and thickness of the magnetized crust

has long been a subject of debate. The depth of the Curie point iso-

therm for magnetite (580



C) is on the order of a few hundred meters

to tens of kilometers below the seafloor (depending on crustal age)

and represents a depth limit for remanent magnetization. Below this

isotherm, all rocks from the oceanic crust could give rise to a remanent

magnetic anomaly signal. However, several decades of magnetic stu-

dies on the oceanic crust showed that magnetization is complex and

varies by several orders of magnitude depending on the structure of

oceanic crust, types of oceanic rocks, and alteration processes.

Magnetic properties of the oceanic crust can be studied by the ana-

lysis of magnetic surveys made using sensing instruments mounted on

submersibles, towed by ships, or incorporated in satellites. Depending

on the distance between the survey and the crustal sources, the focus

of remote magnetic analysis can be at very different scales. Submersi-

ble surveys can be used to map, for example, the meter scale contrast

of magnetization above hydrothermal structures or between flows

along ridge axes. By contrast, satellite surveys will give information

on the properties of wide oceanic areas and the deepest magnetic

sources. In between these extreme scales, sea surface magnetic surveys

adapted to the kilometer scale are the reference. Local information can

also be collected using downhole magnetometer tools in holes drilled

within the crust. When available, the magnetic characterization of drilled

or dredged oceanic rocks is a way to directly access magnetization of

different rock types and is complementary to remote sensing measure-

ments. Ophiolites, ancient ocean-crust fragments on continents, offer a

way to easily access properties of deeper lithologic units. However, the

obduction process probably affects mineralogical properties of rocks in

such material, and it has been observed that less intense magnetizations

are found compared to drilled or dredged oceanic samples. Altogether

the fragmented information collected since the early 60’s allows a global

pattern of magnetization of the oceanic crust to be retreived.

In the following, we present the global structure of the oceanic crust

and its variability, then we review the magnetic properties of the extru-

sive and intrusive basalts, the gabbroic sections, and peridotites.

Global structure of the oceanic crust

In 1972, a model for “typical” oceanic crust structure was derived from

ophiolite observation and seismic velocity analysis. Often referred to

as “Penrose structure,” following the name of the international confer-

ence on ophiolites where it was established, this layered oceanic crust

is as follows: 500 m of basalts, 1500 m of dikes, and 5000 m of gabbros,

which together give an average thickness of 7 km of oceanic crust upon

unaltered mantle. Since the establishment of this basic model, drilling

and dredging show that deep materials, consisting of altered mantle

rocks and gabbros, can be seen at outcrops. This more complex three-

dimensional structure of the crust is due to spatial and temporal varia-

tions of the magmatic and tectonic processes and is more likely to be

596 MAGNETIZATION, OCEANIC CRUST

adapted to slow spreading ridges. Ultramafic rocks such as serpentinized

peridotites, also called serpentinites, are notably reported in these new

crustal models (Cannat et al., 1995). Two “idealized” views of ocean

crust are represented in Figure M116, showing a Penrose “layer cake”

crust and a discontinuous crust with serpentinite emplaced at the surface.

Serpentinized peridotites are altered mantle rocks but will be considered

as part of the magnetic oceanic crust, which is described below.

Basalt layer

The extrusive basalt layer, also referred to as seismic layer 2A (layer 1

being the sediment cover), consists of pillow and sheeted lava erupted

at the ridge axes, pillows being more widely associated with slow

spreading ridges and sheet flows with fast spreading ridges. The sur-

face of flows often consists of a few centimeters of rapidly quenched

glasses. Magnetic properties of extrusive basalts of the oceanic crust

were the subject of numerous studies as a result of the important amount

of dredged and drilled samples made available to the scientific commu-

nity. The magnetic minerals in oceanic basalts are predominantly titano-

magnetites associated with various amounts of titanomaghemites, an

alteration product of titanomagnetites. The primary natural remanent mag-

netization (NRM) of basalts is acquired during the extrusion process and

is a thermal remanent magnetization. Magnetic properties can show sig-

nificant variations between flows depending, in particular, on ironand tita-

nium content but also at the within-flow scale depending on local

crystallization conditions and alteration progression. On average, NRM

of young basalts is in the range of 10–20 A m

1

, but can reach extreme

values on the order of 80 A m

1

or higher for young flows with high

iron content. The Koenigsberger parameter, which is defined as a ratio

of remanent magnetization to induced magnetization, is almost always

greater than one in basalts and can often reach 100 confirming that mag-

netization of the basalt layer is dominantly of remanent origin. This

property favors the magnetic recording of past inversions of the polarity

of Earth’s magnetic field leading to the pattern of lineated magnetic

anomalies. Microscopic observations and Curie temperatures domi-

nantly within 150–250



C for young unoxidized basalts characterize a

primary population of titanomagnetic grains with 60% ulvospinel

(called TM60). In addition to this dominant population within the bulk

part of the lava, fine-grained titanomagnetite with a broader composition

(between 0% and 50% ulvospinel) is observed in the rapidly quenched

glassy part (Zhou et al., 2000). As a consequence, higher Curie tempera-

tures and lower NRMs are commonly found at the pillow margin. How-

ever, magnetic anomalies are caused by basalts rather than the thin glassy

part and the high magnetization of young basalts led some authors dur-

ing the late 1960s to suggest that rocks from only the first few hundred

meters belonging to the extrusive basaltic layers could actually be

responsible for most, if not all, of the magnetic signal observed by Vine

and Mathews (e.g. Klitgord et al., 1975). This led to the first magnetic

models of the oceanic crust as a uniformly highly magnetized basalt

layer. However, extensive magnetic measurements on oceanic rocks

made available, in particular by the Deep Sea Drilling Project (DSDP)

program, in the late 1970s and early 1980s showed that magnetization

of the basaltic extrusive layers was not homogeneous. This is mostly a

consequence of the low temperature alteration of titanomagnetites to

titanomaghemite, caused by circulation of seawater in the rocks pores

and cracks (O'Reilly, 1984). During the process, Curie temperature

progressively increases, whereas NRM decreases typically down to

values on the order of a few A m

1

. A secondary, less intense remanent

magnetization replace part of the original thermal remanent mag-

netization. The magnetic record of Earth’s magnetic field inversions

in the upper oceanic crust is nevertheless well preserved, because this

secondary magnetization remains parallel to the original direction

despite its chemical origin (for a review on this topic see Dunlop and

Ozdemir, 1997). Depending on the alteration environment, secondary

minerals can also be created at the expense of magnetite, iron sulfides

being an example often associated to hydrothermal alteration. The com-

mon occurrence of secondary viscous magnetization contributes to low-

ering of the magnetic signal. These findings corroborate the idea that to

account for the magnetic anomaly signal, deeper seated rocks are more

highly magnetized than first expected.

Dike layer

Below the extrusive basalt layer, dike complexes are the first intrusive

bodies. Oceanic intrusive rocks have been the subject of fewer studies

than basalts due to increasing difficulties in sample accessibility.

Rocks from escarpments, tectonized areas, and ophiolites are thus

overrepresented and might not represent the norm. From what we

know, the sheeted dike layer (also called layer 2B) shows contrasting

magnetizations. The much slower cooling rate of intrusive rocks

results in an initial magnetic population of large multidomain titano-

magnetite grains. Curie temperatures are reported to be quite homoge-

neous and close to 580



C indicating magnetite as the main carrier of

the magnetization. This is the result of widespread oxidation-exsolu-

tion of titanomagnetite grains. This process by which the multidomain

grains exsolved to a solid solution with pure magnetite could take

place during the slow initial cooling (the so-called deuteric oxidation

common in subaerial lavas) but the presence of the characteristic

lamellar structures associated with such a process is not always

observed in submarine dikes. It is likely that low temperature and

hydrothermal alteration make a significant contribution to exsolution

in this context. Hydrothermal alteration also plays an important role

in the loss of iron from grains, recrystallization of magnetite, and

replacement of titanomagnetite crystals by silicate minerals (Smith

and Banerjee, 1986). Typical remanent magnetization in dikes is typi-

cally 10

–1

–10

–2

1

, but stronger magnetizations on the order of

1Am

1

are found in the freshest samples. Koenigsberger ratios are

lower than in basalts but usually above one (ophiolites dikes have lower

Koenigsberger ratio) showing that magnetization of submarine dikes

is for its most part of remanent origin. This remanent magnetization is

mainly acquired during alteration and its relationship with the

geomagnetic field at the time of emplacement is not yet clear.

Gabbros

Gabbros are found below oceanic basalts and are coarse-grained

igneous rocks (seismic layer 3). The dominant magnetic mineral in

gabbros is magnetite, occurring as a primary mineral produced by

the exsolution of large titanomagnetite crystals at high temperature

and as a secondary mineral typically as recrystallization along cracks.

Curie temperatures in gabbros are consequently dominantly found in

the range 550



C–590



C. The intensity of magnetization is quite vari-

able spanning several orders of magnitude and depends on the concen-

tration of magnetic minerals, which is often rather low, giving on

average NRM in the range 10

–1

1

. In a few sites, ferrogabbro

and gabbros rich in olivine were reported and give rather high NRM

Figure M116 Sketch illustrating (a) the uniform oceanic crust

following the model of a Penrose layered crust and (b) an example

of idealized portion of crust with a discontinuity in its center

allowing exposure of ultramafic outcrops (modified after Cannat

et al., 1995). UM, ultramafic rocks; G, Gabbros, D, Dikes, E,

Extrusives.

MAGNETIZATION, OCEANIC CRUST 597

values due to their high concentration of magnetite of primary origin

(in the case of ferrogabbro) or resulting from serpentinization (in the

case of olivine-rich gabbros) (Dick et al., 2000). The Koenigsberger

ratio is often close to one, but usually shows that remanent magnetiza-

tion is dominant over induced magnetization. Remanent magnetization

is often reported to be quite resistant to alternating field demagnetiza-

tion showing good stability, this magnetic behavior, characteristic of

fine-grained population, arises from the subdivision of magnetite in

lamellae structures within exsolved coarse titanomagnetite crystals,

secondary magnetite is often observed as fine-grained population as

well. The contribution of secondary magnetization of chemical origin

(CRM) is difficult to isolate and seems relatively low. Presumably

gabbros can preserve a faithful record of the magnetic field acquired

during the cooling of the units shortly after intrusion at the spreading

axis. Gabbros could contribute to the magnetic anomalies but their

low NRM and depth usually make them marginal contributors. Gab-

bros show important chemical variability and the effect of alteration

on magnetic properties is difficult to single out; studies made on sets

of metamorphosed gabbros show NRMs and Curie temperatures in

the same range as unaltered gabbros (Kent et al., 1978).

Contribution from deeper material: Peridotites

In “typical” Penrose crust, the lower crust is restricted to the gabbroic

layer and no contribution from deeper material is reported. However,

as the occurrence of serpentinized peridotites (also called serpentinites)

within the crust was more and more observed (see Figure M116), their

contribution to magnetic signal became obvious. Unaltered peridotites

in the mantle have a paramagnetic behavior. Magnetization of perido-

tites is acquired during the serpentinization process, which is the reac-

tion of the silicate minerals of peridotites (mostly olivine) with water

producing serpentine and pure magnetite. The amount of magnetite

produced during the serpentinization process is not proportional to the

serpentinization progression resulting in complex iron remobilization

history (Oufi et al., 2002). Even in fully serpentinized peridotites, a

very different behaviors are described with, in particular, Koenigsber-

ger ratios well below or above unity. Curie temperatures are clustering

around 580



C as expected for magnetite. Remanent magnetization is

variable in the range of 10

–1

–10 A m

1

, peak values of 25 A m

1

have been reported, which makes these serpentinite samples as magne-

tized as young basalts (Oufi et al., 2002). From microscopic observa-

tion, magnetite grains can be large and are concentrated along the

veins of serpentines, which is believed to indicate the preferential

direction of fluid circulation. In this case, large grains contribute to a

low NRM and Konigsberger ratio and a strong induced magnetization.

Magnetite is also found as small single-domain grains disseminated in

the serpentinized peridotite matrix, higher NRM and larger Koenigs-

berger ratios are then observed. Serpentinites with such a magnetic

mineralogy could be important contributors to magnetic anomalies.

Nevertheless no clear evidence of serpentinized peridotites recording

a reverse magnetization have yet been established.

Conclusion

The typical range in two of the most used magnetic parameters, NRM

and Curie temperature, are synthesized on Figure M117 for different

rocks within the crust. The effect of serpentinization and low tempera-

ture alteration is represented. As shown and discussed earlier, postem-

placement alteration has a significant effect on magnetic properties of

rocks and contributes to the magnetic heterogeneity of the crust. Due

to their shallow depth, strong magnetic intensity and remanent beha-

vior oceanic basalts are thought to be the most important contributors

to marine magnetic anomalies allowing the establishment of the

sequence of reversals of the Earth’s magnetic field. In addition,

Figure M117 NRM and Curie temperatures frequently observed in oceanic rocks from drilled or dredged samples. Bold lines represent

the average values for “fresh” rocks, the black arrow represents the evolution with serpentinization and small grey arrows show

the evolution of these magnetic properties during low temperature alteration process. Scale bar for photo is 5mm.

598 MAGNETIZATION, OCEANIC CRUST

although this is still a matter of debate, oceanic basalts could provide a

record of the magnetic field intensity variations within polarity intervals,

as suggested by the worldwide occurrence of small scale magnetic var-

iations (called “tiny wiggles”) within the main magnetic anomaly pat-

tern (Cande and Kent, 1992). In any case, it should be mentioned

again that the local crustal architecture is probably complex (Karson

et al., 2002) and is a major factor to take into account in the analysis

of marine magnetic signals.

Julie Carlut and Helene Horen

Bibliography

Cannat, M.C. et al., 1995. Thin crust, ultramafic exposures, and rugged

faulting patterns at the Mid-Atlantic Ridge (22



N–24



N). Geology,

23:149–152.

Cande, S.C., and Kent, D.V., 1992. Ultrahigh resolution marine mag-

netic anomaly profiles—a record of continuous paleointensity var-

iations. Journal of Geophysical Research, 97(B11): 15075–15083.

Dick, H.J.B. et al., 2000. A long in situ section of the lower ocean

crust: results of ODP Leg 176 drilling at the Southwest Indian

Ridge. Earth and Planetary Science Letters, 179:31–51.

Dunlop, D., and Ozdemir, O., 1997. Rock Magnetism. Cambridge:

Cambridge University Press, 573 pp.

Kent, D.V. et al., 1978. Magnetic properties of dredged oceanic gab-

bros and the source of marine magnetic anomalies. Geophysical

Journal of the Royal Astronomical Society, 55: 513–537.

Karson, J.A. et al., 2002. Structure of the uppermost fast spread ocea-

nic crust exposed at the Hess Deep Rift: implications for subaxial

processes at the East Pacific Rise. Geochemistry, Geophysics,

Geosystems, 3: DOI:10.1029/2001GC000155.

Klitgord, K.D. et al., 1975. An analysis of near bottom magnetic

anomalies: seafloor spreading and the magnetized layer. Geophysi-

cal Journal of the Royal Astronomical Society, 43: 387–424.

Morley, L., and La Rochelle, A., 1963. Paleomagnetism and the dating

of geological events. Transactions of the Royal Society of Canada,

1: App 31.

O’Reilly, W., 1984. Rock and Mineral Magnetism. London: Blackie,

220 pp.

Oufi, O. et al., 2002. Magnetic properties of variably serpentinized

abyssal peridotites. Journal of Geophysical Research, 107(B5):

10.1029/2001JB000549.

Smith, G.M., and Banerjee, S.K., 1986. Magnetic structure of the

upper kilometer of the marine crust at deep sea drilling project hole

504B, Eastern Pacific Ocean. Journal of Geophysical Research,

91(B10): 10337–10354.

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

nic ridges. Nature, 199: 947

–949.

Zhou, W. et al., 2000. Variable Ti-content and grain size of titanomag-

netite as a function of cooling rate in very young MORB. Earth

and Planetary Science Letters, 179:9–20.

MAGNETIZATION, PIEZOREMANENCE AND

STRESS DEMAGNETIZATION

With the advent of paleomagnetic studies in the 1940s, scientists

began to wonder how the effect of stress, from burial, folding, etc.,

would influence the magnetic remanence of rocks. By the mid-

1950s, an intense research effort was underway aimed at understand-

ing piezoremanence, which is the remanent magnetization produced

by stress, as well as the opposite effect of stress demagnetization.

The laboratory experiments suggested that the magnetic signals of

rocks were sensitive to stresses typically found around fault zones;

and it was calculated that such stress-induced changes would in turn

modify the local magnetic field. Scientists thought that earthquakes

could be predicted by monitoring magnetic-field variations.

How pressure (stress) affects magnetic remanence

The origin of a spontaneous magnetic remanence is commonly asso-

ciated with electron exchange between iron (or other transition metals

such as Cr, Ni, etc.) atoms. For the iron oxide magnetite (Fe

), the

most abundant magnetic mineral in the Earth’s crust, the exchange is

indirect, passing through an oxygen atom. The spontaneous magnetiza-

tion of magnetite is generated in magnetic lattices established by the

crystallographic arrangement of the atoms. The sum of the magnetic

moments from the lattices comprises the net spontaneous magnetization

of the mineral. The bond lengths and bond angles between iron and oxy-

gen atoms of a particular lattice control the magnetic intensities and

directions of that lattice. Thus, when a stress acts on a material, the bond

lengths and angles change, which will in turn modify the lattices’ mag-

netic moments. This is why the nature of the stress, be it uniaxial, hydro-

static (equal on all sides), compressive, or tensile, can have important

consequences on the magnetic properties of materials. If a material com-

presses isotropically under a pure hydrostatic stress, bond lengths will

decrease, yet bond angles will remain unchanged; whereas, uniaxial or

shear stress will change the bond angles as well as the bond lengths.

Such changes depend on the volumetric compressibility of the magnetic

mineral. For example, the volume of magnetite decreases by about 2%

per gigapascal (GPa) (note that 1 GPa ¼ 10 kilobars (kbar); for the

Earth, 1 kbar corresponds to 3.5 km depth). This is why one needs very

large stresses to significantly modify the electronic configuration and

the magnetic lattice networks.

Size and shape are other factors influencing the magnetic properties

of materials. In extremely small grains, electron exchange exists, but

the spontaneous moment does not remain in a fixed direction within

the crystal due to thermal fluctuations, even at room temperature.

These grains are called superparamagnetic because they have no spon-

taneous moment in the absence of an external magnetic field, yet in the

presence of even a small external field, the randomizing thermal

energy is overcome, and the grains display magnet-like behavior. At

a critical size, about 0.05 mm for needle-shaped magnetite grains, the

spontaneous magnetic moment becomes fixed within the crystal, thus

resembling a dipolar magnet with a positive and a negative pole. Such

grains are called single domain. At even greater volumes, the magnetic

energy of the grain becomes too great and new magnetic domains grow

such that the magnetic vectors of the neighboring domains are oriented

in directions that diminish the overall magnetization of the crystal. These

are called multidomain grains. Multidomain grains often have a net

moment because the sum of the magnetic vectors of all the domains does

not exactly equal zero.

The magnetizations of single- and multidomain grains react differ-

ently to an imposed stress, just as they do with temperature or applied

fields. Hydrostatic stress may cause the net spontaneous moment in a

single-domain grain to reorient to a new position depending on the

grain’s shape, presence of crystal defects, etc. Under subhydrostatic

stress, the magnetization in a single-domain grain will rotate and

change in intensity because different crystallographic axes will experi-

ence different stresses, which will deform the exchange network.

Under any stress condition, the individual domains in a multidomain

grain may grow or shrink to compensate for the deformation, which

will modify the magnetic intensity and direction of the grain. Uniaxial

stresses are more efficient than hydrostatic stresses at causing domain

reorganization.

In sum, two processes contribute to the piezoremanent behavior of

materials, both of which can act independently or simultaneously at

a given stress condition. The first is a microscopic effect that modifies

the electron exchange couple, and the second is more of a macroscopic

effect where the magnetization becomes reoriented within a grain.

Each type of magnetic mineral and the domain states of each mineral

have their own particular piezoremanent and stress demagnetization

MAGNETIZATION, PIEZOREMANENCE AND STRESS DEMAGNETIZATION 599

characteristics. Because rocks are usually composed of a spectrum of

magnetic grain sizes, and sometimes more than one type of magnetic

mineral, in order to understand how and why the magnetic properties

of a rock change in response to an imposed stress, one must know

the composition of the magnetic phases and their size distributions.

Understanding how stress affects a rock must be viewed in a statistical

sense by summing the changes in magnetization of each grain. The

type of imposed stress must also be specified.

Details surrounding magnetic measurements at

high pressure

Magnetic measurements at high pressures are difficult for several rea-

sons. For one, the pressure vessel apparatus must be non-, or only

slightly, magnetic. This is a tall order becaus e most materials strong

enough to transmit high pressures are steel alloys—most of which

are magnetic or interfere with the electronics of the detector. Second,

the sample’s magnetic strength should be detectable by the measuring

system and be significantly higher than the pressure vessel. In order to

maximize the magnetic signal, one wants as many magnetic grains as

possible in the sample. It follows that larger samples are more mag-

netic, and more easily measured than smaller ones—not because the

density of magnetic grains is greater, but because of a greater number

(or mass) of magnetic grains. As pressure equals force divided by area,

a trade-off exists between pressure and sample size, which explains

why most experimental work on this problem has focused on large,

rock samples. The majority of experiments have been performed using

uniaxial stresses and pressures of less than 1 GPa. Indirect measurements

of the magnetic states of minerals at much higher pressure cells have

been employed since the 1960s using Mössbauer spectroscopy, which

measures magnetic interactions between electrons and the nucleus.

Although Mössbauer data yields important information about the exis-

tence and type of magnetic network in materials, it does not quantify

the strength or direction of the moment or their stress dependencies

and thus is beyond the scope of this summary. Direct measurements of

the full magnetic vector to significantly higher pressures (to 40 GPa)

are now becoming possible with the development of diamond anvil cell

technology. This new method also allows one to experiment on magnetic

minerals whose composition and domain state are well characterized.

An important observation of the high-pressure experiments pre-

formed under uniaxial stress is that magnetic properties change in rela-

tion to the maximum stress axis direction. For stress demagnetization

(i.e., stress applied in the absence of an applied magnetic field), the

magnetization intensity decreases faster parallel to the maximum stress

direction rather than perpendicular to it. When pressure is applied in

the presence of a magnetic field, magnetic vectors in the grains rotate

toward the direction perpendicular to the maximum stress direction.

Both magnetic susceptibility and magnetization increase perpendicular

to the maximum stress direction and decrease parallel to the maximum

stress direction (Figure M118). This has important implications

because several experimental apparatuses can only measure magnetic

parameters in one direction. Obviously this will lead to a bias of the

interpretations of the data.

A special case of stress demagnetization and piezoremanent magne-

tization is called shock demagnetization and shock magnetization. The

term shock is added when the application and removal of a stress

occurs in a brief instant in time, such as during a meteorite impact

or by firing a projectile at a magnetic sample. To date no study has

established that either stress demagnetization or piezoremanence is a

time-dependent process, so the difference in terminology should be

considered semantic.

Stress demagnetization

Despite the experimental restrictions, several things are known about

the magnetic properties of rocks and minerals under stress. The effect

of stress in the absence of an external magnetic field has a permanent

demagnetizing effect on a rock composed of several magnetic grains.

In this case, permanent means that once the pressure is released, the

magnetic moment does not grow back to its original, precompressed

value because the magnetic vector directions of the grains have been

randomized, thus lowering the net magnetization of the rock. It does

not mean that the individual magnetic grains have lost their capacity

to be magnets. This makes stress demagnetization similar to alternat-

ing field demagnetization. Place a rock in a decaying alternating field

in a null or very weak external field and the rock will demagnetize.

However, the magnetic intensity of a rock placed in a decaying alternat-

ing field in the presence of an external field will often increase to levels

Figure M118 Relative change in magnetic susceptibility (w) measured perpendicular and parallel to the maximum compression

direction in two titanomagnetite samples (Curie temperatures of 250



C and 200



C, respectively) (after Kean et al., 1976).

600 MAGNETIZATION, PIEZOREMANENCE AND STRESS DEMAGNETIZATION

higher than its starting intensity. The same is true for pressure (stress)

demagnetization and its converse, piezoremanent magnetization.

Figure M119 shows a good example of the differences between hydro-

static and uniaxial stress demagnetization of a diabase rock containing

large (20–300 mm) titanomagnetite (Curie temperature ¼ 535



grains. The pressures employed in these experiments are relevant

to the Earth’s crust, as 100 MPa (which equals 1 kbar or about 3.5 km

depth) is well within the seismogenic zone, i.e., the upper region of the

Earth’s crust where earthquakes are generated. One quickly sees that

uniaxial stress demagnetizes this material about two times faster than

hydrostatic stress; e.g., at 100 MPa, hydrostatic and uniaxial pressures

demagnetize roughly 20% and 40% of the original, noncompressed

magnetization intensity, respectively. This is true whether the material

is under pressure, or if it has been decompressed. Upon decompression,

a demagnetization effect also exists, which is greater under uniaxial than

hydrostatic stress. It should also be noted that during pressure cycling,

once a rock has been compressed and then decompressed from a given

pressure, it resists further stress demagnetization up until the maximum

pressure that the rock has previously experienced. Once that pressure is

exceeded, the rock continues demagnetizing in a manner coherent with

the previously defined path. The pressure history dependence can be

likened to a work or magnetic hardening effect owing to the strain state

of the grains. Multidomain grains are more sensitive to stress than sin-

gle-domain grains because the magnetic remanence of multidomain

grains is controlled by domain wall migration, which is highly sensitive

to dislocations and other imperfections in the crystal.

Piezoremanent magnetization

Imagine a rock that has acquired its magnetization by cooling in the

presence of the Earth’s magnetic field. The magnetic intensity of the

rock will be proportional to the strength of the field in which it cooled,

the number and type of magnetic grains, their sizes and shapes, and the

way the grains are geometrically distributed in space. For a rock whose

grains are randomly distributed (no preferred orientation), the mag-

netic moment of each nonspherical grain will tend to lie parallel to

its long-axis, favoring the long-axis direction pointing toward the mag-

netic field direction. If a relatively low magnetic field is then applied,

say three times that of the present Earth’s field, the rock’s magnetiza-

tion will slightly increase, proportional to the amount of magnetic

material in the rock (as defined by its magnetic susceptibility). After

the field is removed, the magnetization will decay reversibly back to

its original value. When relatively high fields, say 100 times the

Earth’s field, are applied, the magnetization directions of the grains

rotate parallel to the field, leading to a greater net magnetization of

the rock. Upon removal of the field, the rock’s magnetic intensity will

decrease, but will still remain above the value prior to the application

of the strong field. This is due to an irreversible reorganization of the

individual moments in each grain. As successively higher fields are

applied and then removed, the magnetic moments of the grains

become more and more aligned parallel to one another, again due to

irreversible rotation, leading to greater and greater magnetic intensi-

ties. At some applied field value, the moments of all the grains become

aligned to the extent that successively higher fields can no longer

cause irreversible rotations and the rock’s magnetic intensity remains

constant. This point is called the saturation remanence of isothermal

remanent magnetization (SIRM) and the plot of magnetic intensity as

a function of applied field is an isothermal remanent magnetization

(IRM) curve (Figure M120).

When a magnetic material is subject to stress in the presence of a

magnetic field, the material will acquire a remanence proportional to

the strength of the field and the level and type of stress. Figure M120

shows four IRM curves obtained from the same titanium-free, multido-

main magnetite sample (a) in the stress-free state (had never been sub-

jected to any pressure), (b) when the sample was compressed and held

at a hydrostatic pressure of 3.13 GPa, (c) when the sample was further

compressed and held at a hydrostatic pressure of 5.96 GPa, and (d)

after decompression (P ¼ 0) from P ¼ 5.96 GPa. Piezoremanence

refers to the phenomenon where the application of stress in the pre-

sence of a magnetic field increases the material’s magnetization above

the level that it would have acquired if the same field had been applied

in the absence of stress; e.g., by definition, a piezoremanent magneti-

zation lies anywhere in the gray region above the IRM curve (a) in

Figure M120. For this multidomain magnetite example, one observes

that the magnetization gained at any applied field increases with

increasing pressure (Figure M120, curves b and c). This means that

pressure (stress) enhances irreversible rotations of the grains’ moments,

by domain reorganization, more than an applied field does in the

absence of stress. After pressure is released, the domain structures of

the grains are even more disposed to irreversible rotation than before

(Figure M120, curve d).

The sequence that stress and magnetic fields are applied in will

modify a rock’s magnetism in different ways and is again dependent

on the species of magnetic mineral and its domain state. For the case

in Figure M120, one observes that the net magnetization of a multido-

main magnetite sample having been (1) brought up in pressure,

(2) subject to an applied field, and then (3) removed from the field

(curves b and c) is less than having been (1) brought up in pressure,

(2) removed from pressure, (3) subject to an applied field, and then

Figure M119 An example of the differences between hydrostatic

and uniaxial stress demagnetization for a diabase sample

containing multidomain titanomagnetite (after Martin and

Noel, 1988).

MAGNETIZATION, PIEZOREMANENCE AND STRESS DEMAGNETIZATION 601

(4) removed from the field (curve d). The opposite is generally true for

single-domain magnetite grains. The reason for this is partly due to

the way domains facilitate irreversible rotation. Other contributions

to irreversibility may come from grain-size reduction (e.g., breaking

multidomain grains into single-domain grains) or a permanent modifi-

cation of the exchange couple. Note that the magnetization of a sample

(1) brought up in pressure, (2) subject to an applied field, (3) removed

from the field and then (4) removed from pressure may or may not

exhibit piezoremanence. This depends on whether compressive stress

produces more irreversible rotation than the demagnetization effect

that accompanies decompression.

Piezoremanence and stress demagnetization in nature

On Earth, piezoremanent magnetization and stress demagnetization are

relevant anywhere transient or permanent stresses provoke a measur-

able change either in a short-term sense (e.g., during an earthquake)

or in a long-term sense (e.g., during the build up or relaxation of stress

in the crust due to tectonic motion). Mathematical models of fault

motion during earthquakes estimate local deviations in the magnetic

field intensity and may reach 1–10 nT (the present Earth’s field inten-

sity ranges from 30000 to 60000 nT), which is well within the measur-

able limit of about 0.1 nT. The size of the seismomagnetic effect

depends on several parameters including: magnetic mineralogy,

domain state, magnetization direction and intensity of the country

rock, direction and intensity of the Earth’s field, strike and dip of the

fault, sense and amount of motion on the fault, and earthquake magni-

tude and depth. An important unknown is the stress history of the rock.

As shown above in Figure M119, rocks undergo a magnetic hardening

during stress demagnetization. So, for a significant change in the

rock’s magnetization to occur, the rock must be stressed to higher

levels than it has previously seen. This could be why piezoremanent

effects during an earthquake have never been witnessed. On the other

hand, it is difficult to set up and maintain an array of magnetometers

over very long time-periods. As earthquakes cannot be predicted, the

chances of being at the right place at the right time are extremely low.

Long-term changes in the local magnetic field have been documen-

ted around faults and are thought to coincide with nonseismic stress

loading due to plate motion. Moreover, long-term changes in the local

magnetic environment were observed when large reservoirs were filled

with water. The changes were attributed to ground loading and the piezo-

magnetic effect. Magnetic field variations around volcanoes have been

observed during or preceding volcanic eruptions. Although the changes

were originally attributed to stress-field changes during the charge or

withdraw of magma, closer examination of the data suggests that the

magnetic variations were induced by fluid circulation. A similar debate

exists for certain subduction zones that exhibit long wavelength

magnetic anomalies. Subduction of cold lithosphere depresses the

Curie isotherm to depths down to 100 km. Thus, magnetic minerals

can be displaced to, or form at, great depths and still be above the

Curie isotherm. Note that the Curie temperature of titanomagnetite

increases by 2



kbar

–1

. Some workers attribute the anomalies to

piezoremanence while others believe it is related to fluid flow in

the subduction zone.

Meteorite craters are among the most likely place to find the effects

of piezoremanence or stress demagnetization. The presence of shatter

cones, pseudotachylite, microscopic textures such as planar de-

formation features, and large volumes of melted rocks, all suggest that

pressures during impact range from about 5 to more than 30 GPa, for

reasonably sized (>20 km diameter) craters. However, finding clear-

cut evidence that the magnetism of the country rocks has been affected

by shock is not so easy because most craters are old (>1 Ga) and have

been severely eroded, deeply buried, or the thermal effects of the

impact have not been eroded deeply enough to expose the shocked

rocks. Moreover, after several hundreds of millions of years, the origi-

nal magnetic mineralogy of the shocked rocks may become altered

by oxidation, or new magnetic minerals may grow. However, the fact

that aeromagnetic anomalies are characteristic features of meteorite

impacts, there must be some long-term memory of the country rock’s

magnetization. Some scientists have found that the shocked rocks are

remagnetized in the direction of the Earth’s field at the time of impact.

If true, one can perform an impact test by collecting samples along a

Figure M120 Isothermal remanent magnetization (IRM) curves for the same sample of titanium-free, multidomain magnetite at

(a) initial (zero pressure), (b) and (c) under hydrostatic pressures, and (d) at zero pressure after decompressing from 5.96 GPa.

Any point lying in the gray area is a piezoremanent magnetization. (S. Gilder, M. LeGoff, J.C. Chervin and J. Peyronneau,

unpublished data).

602 MAGNETIZATION, PIEZOREMANENCE AND STRESS DEMAGNETIZATION

radial transect away from the center. The remagnetization direction

should be most prevalent near the center and eventually die out with

distance.

An intriguing case for shock demagnetization has been proposed for

Mars. The southern hemisphere of the Martian crust is highly magnetic

except for two very large (>500 km diameter) circular regions that

have significantly lower magnetic intensities. Some scientists believe

these areas of low magnetic intensity are due to stress demagnetization

that occurred during meteorite impacts. Although the magnetic mineral

in Martian rocks is presently thought to be magnetite, little is known

about the piezomagnetic and stress demagnetization behavior of this

substance at the cold (9



Cto–128



C) temperatures of the red

planet’s surface. Despite some 50 years of research, our knowledge

of piezoremanence and stress demagnetization remains limited. A

renaissance in this research field, partially due to recent space explora-

tion, will surely lead to exciting discoveries in the future.

Stuart Alan Gilder

Bibliography

Kean, W., Day, R., Fuller M., and Schmidt, V., 1976. The effect of uni-

axial compression on the initial susceptibility of rocks as a function

of grain size and composition of their constituent titanomagnetites.

Journal of Geophysical Research, 81: 861–872.

Martin, R., and Noel, J., 1988. The influence of stress path on thermore-

manent magnetization. Geophysical Research Letters, 15:507–510.

Cross-references

Magnetic Remanence, Anisotropy

Magnetization, Thermoremanent (TRM)

MAGNETIZATION, REMANENT, AMBIENT

TEMPERATURE AND BURIAL DEPTH FROM

DYKE CONTACT ZONES

Introduction

Nagata (1961, p. 320) proposed using the directions of remanent

magnetization in the contact zone of an igneous intrusion as a

geothermometer. Carslaw and Jaeger (1959) and Jaeger (1964) applied

heat conduction theory to calculate the variation in maximum reheat-

ing temperature with distance from a cooling intrusion of simple geo-

metry. Schwarz (1976, 1977) carried out the first study in which the

paleomagnetism across a dike contact and theoretical maximum tem-

perature curves were used to estimate the ambient temperature of the

host rocks and the depth of burial of the present erosion surface at

the time of dike intrusion.

In principle, the depth of burial method can be applied to the

contacts of a variety of intrusions with different shapes. However, cal-

culating the maximum reheating temperature in the vicinity of an

intrusion with complicated geometry is difficult. To date, depth of

burial studies that use remanent magnetization have usually been con-

fined to dike contacts. The case of a dike contact is used to illustrate

the method in this article.

Magnetic overprinting in the vicinity of a dike

The depth of burial method is based on the thermal overprinting of

remanent magnetization in host rocks near dike contacts. It has been

reviewed by Buchan and Schwarz (1987).

When a dike intrudes, temperatures in adjacent host rocks rise above

the highest magnetic unblocking temperatures (T

). During subse-

quent cooling in the Earth’s magnetic field, the dike and this adjacent

“baked” zone acquire magnetic remanences with similar directions

(Figure M121). Far from the dike contact, in the “unbaked” zone, peak

temperatures are insufficient to thermally reset even the lowest block-

ing temperature component (see Baked contact test). Between the

baked and unbaked zones lies a region of hybrid magnetization where

reheating is only sufficient to result in thermal resetting of the lower

blocking temperature portion of the host rock magnetization. If, in this

“hybrid” zone,

the older and younger components of magnetization

can be separated on the basis of their blocking temperature spectra,

then the maximum temperature (T

max

) reached following dike intru-

sion can be determined as a function of distance from the contact

(Schwarz, 1977). This maximum temperature may be corrected for

the effects of magnetic viscosity during prolonged heating in the field.

Using heat conduction theory (Carslaw and Jaeger, 1959; Jaeger,

1964), the ambient temperature of the host rock just prior to dike intru-

sion can then be calculated. Depth of burial of the present erosion

surface is determined by dividing the ambient temperature of the host

rock by the estimated geothermal gradient at the time of intrusion.

It should be noted that the study of the magnetic hybrid zone

according to the procedure of Schwarz (1977) also yields the most

rigorous type of baked contact test (the baked contact profile test), a

field test for establishing the primary nature of the remanent magneti-

zation of an intrusion (see Baked contact test).

Assumptions

Several assumptions are made in the following discussion of the depth

of burial method applied to the remanent magnetization across a dike

contact (Buchan and Schwarz, 1987).

1. The dike can be approximated by an infinite sheet.

2. The dike was formed by a single magma pulse of short duration.

3. Heat transfer occurs solely by conduction.

4. Dike intrusion results in a (partial) thermal resetting of the rema-

nence of the host rock, without chemical changes.

5. Magnetic mineralogy in the hybrid zone is relatively simple, so that

the effects of magnetic viscosity can be estimated from published

time-temperature curves.

Complicating factors such as prolonged duration of magma flow in

the dike, multiple magma pulses, groundwater, volatiles escaping from

the dike, and chemical resetting of remanence in the hybrid zone are

Figure M121 Directions of remanent magnetization with distance

from the contact of an igneous intrusion. It is assumed that the

magnetic mineralogy of the host rock is similar throughout the

profile and that overprinting in the baked and hybrid zones is a

thermoremanent magnetization.

In some publications the baked, hybrid, and unbaked zones are referred to as the

contact, hybrid, and host (magnetization) zones, respectively.

MAGNETIZATION, REMANENT, AMBIENT TEMPERATURE AND BURIAL DEPTH FROM DYKE CONTACT ZONES 603

usually difficult to take into account (see Buchan et al., 1980; Delaney,

1982, 1987; Delaney and Pollard, 1982; Schwarz and Buchan, 1989).

They are not addressed in this article.

Sampling

Oriented paleomagnetic samples are collected along a continuous pro-

file that includes the dike, and the baked, hybrid, and unbaked zones

of the host rocks (e.g., Figure M122). The horizontal distance of each

sample from the contact is recorded, along with the horizontal width of

the dike. If the dike is tilted from vertical, the perpendicular distance of

each sample from the contact and the thickness of the dike should be

calculated.

Determining maximum reheating temperature (T

max

)

in the hybrid zone

Each hybrid sample is thermally demagnetized in a stepwise fashion.

In general, purely viscous components are eliminated at low tempera-

tures to reveal magnetization that is the sum of a host component and a

component acquired at the time of emplacement of the igneous unit

(Figure M121). Upon thermal demagnetization to higher temperatures,

the remanence direction of each hybrid sample moves progressively

along a great circle path to the host direction (e.g., Figure M123). This

reflects the fact that the T

spectra of the overprint and host mag-

netization components are essentially discrete, with the overprint

spectrum occupying a range immediately below that of the host

component.

max

, the maximum temperature attained at a given locality in the

hybrid zone following dike intrusion is given by the highest T

the overprint component. It can most easily be obtained using an

orthogonal component (or Zijderveld) plot of the horizontal and verti-

cal projection of the thermal demagnetization data (e.g., Figure M124).

max

is given by the temperature at which the straight-line segments

through the overprint and host components intersect (Dunlop, 1979;

McClelland Brown, 1982; Schwarz and Buchan, 1989; Dunlop and

Özdemir, 1997). In the example that is shown in Figure M124 the inter-

sections are fairly sharp so that T

max

values are readily determined.

Note that the interpretation of more complicated orthogonal compo-

nent plots involving multiple magnetic minerals and chemical

overprinting are discussed in Schwarz and Buchan (1989).

The T

max

values determined from the orthogonal component plots of

Figure M124 decrease progressively across the hybrid zone with

increasing distance from the dike contact as expected for thermal over-

printing due to the dike emplacement.

Correcting T

max

for magnetic viscosity

The maximum blocking temperatures (T

max

) that are reset in a given

thermal event are dependent upon the length of time over which the

elevated temperatures are maintained (Dodson, 1973; Pullaiah et al.,

1975). Therefore, T

max

must be corrected using published time-

temperature curves for the appropriate magnetic mineral (e.g., Pullaiah

et al., 1975).

Determining ambient temperature of host rock at time

of dike emplacement (T

amb

)

The maximum temperature (T

max

) attained at a particular locality in the

host rock following dike intrusion equals the sum of the maximum

temperature increase (DT

max

) due to heat from the dike and the ambi-

ent temperature (T

amb

) of the present surface just before the dike was

emplaced.

Therefore, in the hybrid zone, where T

max

falls within the overall

unblocking spectrum, T

amb

is calculated from the equation:

amb

¼ T

max

 DT

max

(Eq. 1)

Different heat conduction models can be utilized to determine the

maximum increase in temperature (DT

max

) in the contact zone of a

dike (Carslaw and Jaeger, 1959; Jaeger, 1964; Delaney, 1987).

The simplest model that accounts for the latent heat of crystalliza-

tion of the dike magma is one in which the heat released during soli-

dification is added to the initial magma temperature as an equivalent

temperature increase (Jaeger, 1964). This model gives a reasonable

approximation to more rigorous models for distances greater than a

quarter of a dike width from the contact, and has been employed

in a number of studies of dike contacts where the hybrid zone is rela-

tively far from the contact (e.g., Schwarz, 1977; Buchan and Schwarz,

1981; Schwarz and Buchan, 1982; Schwarz et al., 1985; Hyodo et al.,

1993; Oveisy, 1998).

Using this simple model, the maximum temperature increase at a

distance x from the center of a dike of thickness 2d is calculated using

the following equations from Jaeger (1964), modified slightly to allow

for nonzero T

amb

max

¼½V þðL=cÞðT

max

 DT

max

Þ



½erffðx þ dÞ=2dt

1=2

gerffðx  dÞ=2dt

1=2

g

(Eq. 2)

where t ¼ kt/d

, k ¼ k/(rc), and erf (u) is the error function.

V and L are respectively the intrusion temperature and latent heat of

crystallization of the dike magma, k, k, r, and c are respectively the

diffusivity, conductivity, density, and specific heat of the host rock,

t is the time elapsed since intrusion of the dike, and t is dimensionless

time.

Figure M122 Sampling profile perpendicular to contact of a

ca. 2200–2100 Ma Indin diabase dike of the Slave Province of the

Canadian Shield (modified after Schwarz et al., 1985). The dike

intrudes Archean metavolcanic rocks.

604 MAGNETIZATION, REMANENT, AMBIENT TEMPERATURE AND BURIAL DEPTH FROM DYKE CONTACT ZONES

max

is determined from the equation:

max

¼ðx=dÞ= ln½ðx þ dÞ=ðx  dÞ (Eq. 3)

which is satisfied when the temperature reaches its maximum value,

max

In this model, the required thermal parameters are assumed to be

constant. Typical values are: V ¼ 1150



CandL ¼ 3.77  10

Jkg

1

for mafic dike rocks; c ¼ 1.26  10

Jkg

1



1

for mafic host rocks;

and c ¼ 1.09  10

Jkg

1



1

for granitic host rocks (e.g., Touloukian

et al., 1981).

The simple heat conduction model described above is not valid

within a quarter of a dike width of the contact (Jaeger, 1964), because

it does not adequately account for the release of the latent heat of

crystallization during cooling of the dike or the temperature depen-

dence of thermal conductivity and diffusivity (Delaney, 1987). In some

magnetic studies, hybrid samples have been obtained close to the dike

(Buchan et al., 1980; Symons et al., 1980; McClelland Brown, 1981),

so that the simple model is inappropriate. The authors in these studies

applied more rigorous analytical or numerical models described by

Carslaw and Jaeger (1959).

The most sophisticated analysis of heat conduction models for a

dike cooling by conduction has been described by Delaney (1987,

1988). He demonstrated the advantage of numerical solutions over

analytical solutions, especially at and close to the dike contact. Delaney

(1988) published computer programs to calculate the temperature in

a dike and host rocks as they cool conductively, in which the

temperature dependence of thermal properties and the latent heat of

crystallization are taken into account. These programs have been

applied to the case of remanent magnetization in a dike contact by

Adam (1990).

Finally it should be noted that T

amb

may be determined in one of

two ways. It can be calculated from individual T

max

values (e.g.,

Schwarz and Buchan, 1982). Alternatively, when there is a suffi-

ciently wide hybrid zone, T

amb

can be determined by comparing the

whole experimentally determined profile of T

max

values with the the-

oretical T

max

curves of maximum temperature based on conductive

heating of the dike (McClelland Brown, 1981; Buchan and Schwarz,

1987).

Determining depth of burial of the present erosion

surface at the time of dike emplacement

The depth of burial of the present erosion surface can be determined

by dividing T

amb

(determined from Eq. (1)) by the paleogeothermal

gradient for the sampling area at the time of dike intrusion. To estimate

the paleogeothermal gradient, the present-day geothermal gradient

must be corrected for the decrease in heat generation resulting from

the decay of radiogenic isotopes since the time of dike intrusion

(e.g., Jessop and Lewis, 1978), and in some areas for the effects of

Figure M123 Thermal demagnetization characteristics of selected samples from the Indin dike profile of Figure M122 (after Schwarz

et al., 1985). Distance to the contact is given for each sample. Directions are plotted on equal area nets with closed (open) symbols

indicating positive (negative) inclinations. Heat-treatment temperatures are indicated in degree Celsius. Data points that represent the

great circle between dike and host magne tization directions are connected by a dashed line. T

max

values, estimated from the

experimental data before correction for magnetic viscosity, are given for hybrid zone and host zone samples.

MAGNETIZATION, REMANENT, AMBIENT TEMPERATURE AND BURIAL DEPTH FROM DYKE CONTACT ZONES 605