INTRODUCTION

Materials acquire induced magnetization (Amperes per meter) when placed in an external magnetic field (Teslas). This induced magnetization is temporary and vanishes when the external field is removed (reduced to zero). Under some circumstances materials can become permanently magnetized; the magnetization remains even when the external field is no longer present. These magnetizations may remain unchanged over geologic time and are called remnant magnetizations. There are several important kinds but all depend on atom-to-atom coupling of electron spins in ferromagnetic materials.

THERMAL REMANENT MAGNETIZATION (TRM)

In 1853, on a field trip to Mount Vesuvius, Melloni collected a set of samples of recently cooled lava. He was careful to record the orientation of the samples (which way was up, which way was north). Back in the lab he surveyed the magnetic field emanating from the samples. In this way he found that the dipole axes of his samples were parallel to the magnetic field vector at the field sites. I’m not certain if he retested the samples to see if these magnetizations changed over time. Ideally one uses electrical coils to cancel the field of the earth in the lab during the measurements and thus eliminate induced magnetization (visit Dr. Clyde’s lab in the basement).

Subsequent work with many recent lava flows confirms Melloni’s discovery. As lavas cool they become solid and, at temperatures below about 600 oC (for magnetite) they become magnetized parallel to the ambient (external) field. The magnetization of an igneous crystal becomes stable when the ratio of crystal volume to temperature exceeds some critical or threshold value. Thus the big magnetite crystals "lock in" the field at higher temperatures than small grains. We can actually estimate the magnetite grain size distribution from the graph of magnetization versus temperature.

In detail the magnetic stability depends on crystal size in a complex way. In tiny crystals the thermal energy is so large that the magnetization never locks in. In large crystals the volume is split into several "domains", each with its own direction of magnetization. As the external field changes over time the domains grow and shrink, leading to a time-varying, unstable magnetization. These large crystals typically occur in plutonic rock and thus in continental rocks, as common on the continents, often have little or no significant remnant magnetization.

Crystals of intermediate volume are "single domain" and hold their TRM very well. Such crystals are common in basalt flows and thus remnant magnetization in the ocean crust is very stable (but see below). TRM can be changed or even eliminated by reheating of the magnetized crystals.

CHEMICAL REMANENT MAGNETIZATION (CRM)

Sedimentary rocks may become magnetized if a cement such as hematite grows in the pores. Here again the magnetization locks in the ambient field direction when the volume to temperature ratio exceeds a threshold value. In TRM the ratio rises because the temperature decreases (the grains are already fully grown) and the age of the magnetization is almost the same as that of the rock. In the case of CRM, however, the ratio rises as the crystal grows at constant temperature and the magnetization may be millions of years younger than the time of deposition of the sediments.

Many studies have shown that the magnetization of deep-sea basalts decreases to about ten percent of its initial value within 10^4 to 10^6 years of eruption because of hydrothermal alteration. The process involves a complex series of crystal decay and growth. The net results seem to be a weak but stable magnetization in the original direction.

DETRITAL REMANENT MAGNETIZATION (DRM)

This kind of magnetization occurs mainly in fine-grained sediments and sedimentary rocks. These sediments, made of clay flakes and silt chips, are often of very high initial porosity (house of cards model) and may contain grains of magnetite and other magnetic minerals weathered out of the source rocks. Very small magnetite crystals are too small to hold a stable magnetization (volume/temperature effect) whereas large ones are oriented by gravity and water flow. Grains of intermediate size may actually rotate like little compass needles and thus record the ambient field direction. Given the house of cards fabric of fine-grained sediments it’s not surprising that the DRM is altered as the sediments eventually are compacted; this commonly leads to estimates of inclination that are too low.

In some limestones and deep-sea "red clays" there are little if any magnetic grains except those grown by magnetotactic bacteria, birds, reptiles, mammals (including humans).

ISOTHERMAL REMANENT MAGNETIZATION (IRM)

If ferromagnetic minerals are exposed to extremely strong magnetic fields they may become magnetically saturated such that atom-to-atom electron spin alignment may continue even when the external field returns to zero. The first hint of this behavior was the discovery of magnetically reversed rocks (or at least places where the compass pointed south) by Joa de Castro in India almost five hundred years ago. Around 1800 Alexander von Humboldt showed that lightning strikes could produce this effect as these magnetically reversed rocks preferentially occur at bedrock outcrops at the tops of high cliffs (These are not the reversed rocks that cause the magnetic stripes of the ocean floor).

VISCOUS REMANENT MAGNETIZATION (VRM)

This magnetization is the cumulative effect of thermal vibrations acquired over time at low temperatures (that is, well below the Curie temperature). VRM is parallel to the external field and grows more rapidly the higher the temperature. The polarity of some samples may reverse within a year just sitting around the laboratory! Clearly this is an insidious complicating factor in studying magnetic rocks and brings up the topic of stability of remanence of any kind.

STABILITY TESTS

There are numerous laboratory tests of the stability of remnant magnetization. They all are based on the supposition that if the NRM is easy to alter or destroy in the lab then it has a good chance of being unstable over geologic time. Conversely, if it’s hard to change, perhaps it really has existed unchanged for millions of years.

Thermal demagnetization tests determine at what temperature the magnetization vector is eliminated or significantly reduced. If the remanence can be altered at low temperatures it suggests lack of stability as the rocks may well have been subjected to such temperatures in the course of their long history.

Both AC and DC magnetic fields are used in lab tests for stability. Again a linkage is postulated between "hard to destroy" and "stable". Chemical leaching is also used as is crystal-by-crystal destruction by lasers.

We should be a little skeptical about such short time lab tests as none can really account for possible changes over geologic time. For that reason a whole set of classical geological tests have been devised.

Perhaps the most basic is the "consistency test". Here we measure the NRM of a variety of different ideologies of the same age and in known relative locations. If the NRM of all sites and ideologies is consistent with the same axial dipole model then the remanence could be stable. If sites of different ages can be synthesized into a coherent story, so much the better. The basic assumption, of course, is that initially similar TRM, CRM and DRM directions would change differently over time unless all were stable.

The "baked contact test" compares the NRM of an intrusive body (say a sill) with that of country rock near the sill (in the baked contact zone) and far away. If the NRM of the sill is the same as that in the upper and lower baked zones then perhaps it is stable. If the directions are different then at least one of the lithologies does not have stable NRM.

If all the clasts in a conglomerate have the same NRM then we assume that the NRM is unstable, both in the conglomerate and in the source rocks. If each clast has its own direction then perhaps the NRM is stable.

The fold test uses the relationship of NRM direction to bedding in folded strata. If the NRM direction is independent of orientation of the beds the NRM is post-folding and presumably not stable (at least with respect of time of deposition). If the NRM makes the same angle with bedding at all sites than it is pre-folding in age and, hopefully, stable since the strata were laid down.

MAGNITUDE OF NRM VECTOR

Magnetic remanence is a vector quantity with both direction and magnitude. Essentially all we have looked at so far is stability of NRM and recovery of pale magnetic directions. Why haven’t I said much about magnitude?

Directions are easy to measure and somewhat immune to changes in the rocks over time as they depend on ratios of vertical and horizontal magnetizations. In the absence of strong fabric effects we would expect both components to change in the same way.

Original magnitudes of the magnetization vector are very hard to estimate. Basically one has to subtract out the effect of all the changes that occurred over millions of years or more. The basic strategy is to try to totally demagnetize the sample without changing the mineralogy in any significant way. For example we might thermally demagnetize a basalt sample in a neutral atmosphere of Nitrogen. Then , step-by-step, we remagnetize it up to the point where the original NRM is recovered. The lab field is assumed equal to the original ambient field. But perhaps the rock had a more complex history inadequately mimicked by our experiment. We’ll look at some results later.