Geomagnetism

Because these fluctuations in the geomagnetic field do not follow a simple, mathematically predictable cycle, records of past secular variation must be compiled from sources such as historic observations of the field (Barraclough 1995; Malin and Bullard 1981), archaeomagnetic measurements of archaeological features (Bucur 1994; Clark et al. 1988; Eighmy 1991; Kovacheva 1980, 1997; Kovacheva et al. 1998), paleomagnetic measurements of lake sediments (Lund and Banerjee 1985) or lava flows (Champion 1980; Doell and Cox 1965; Hagstrum and Champion 2002), or some combination of the above (Lund 1996; Thompson 1982). Furthermore, the spatial variations created by the nondipole components necessitate the development of separate records for each archaeomagnetic region (Tarling 1983). Numerous researchers have addressed the question of the optimal size for archaeomagnetic regions (Noël and Batt 1990; Shuey et al. 1970; Sternberg 1982, 1997), and have reached the general consensus that archaeomagnetic records are valid over areas of approximately 1000 km in diameter. Areas larger than this will incorporate extensive directional scattering, which will introduce significant errors into the compiled secular variation record (Batt 1997:153; Noël and Batt 1990: Figure 3, 4; Sternberg 1997:326).

Fired, stationary clay features such as hearths, earth ovens, or buildings serve as accurate paleomagnetic recorders due to the properties of the magnetic minerals typically present in clay. When clay-rich features are heated close to or above mineralogically specific temperatures (i.e., Curie temperatures; Tcmagnetite = 580° C, Tchematite = 680° C) and then allowed to cool back to ambient temperatures, they acquire a magnetization that parallels the prevailing magnetic field (e.g. the geomagnetic field). A feature will retain this magnetization unless it is reheated to a similar temperature, at which point a new magnetization will be acquired. Thus, the magnetization that is measured in the lab is assumed to have been acquired when the feature was last fired, which in turn is related to a particular cultural event (e.g., remodeling, burning, or abandonment of a house or site).

The geomagnetic field can be described at any point on the earth’s surface by a vector that indicates the direction and strength of the field. The field’s direction is typically discussed in terms of its inclination, or dip, and declination, or strike.

By recording these measurements for numerous spots around the globe, researchers can determine the general structure of the global field (Butler 1992; Sternberg 1982, 1990; Tarling 1983). Roughly 90 percent of the measured field can be accounted for by a simple dipole (similar to a bar magnet) placed at the center of the earth and inclined from the rotation axis by about 11.5 degrees (Butler 1992, Tarling 1983). This is the inclined geocentric dipole model. When the field predicted by this model is subtracted from the observed geomagnetic field, a heterogeneous residual field remains. This is called the nondipole field, and it’s concentrated into about six or seven continent-sized features. These features are subject to growth, decay and movement, and these changes occur over periods of anywhere between a couple decades to several millennia. These, in combination with more uniform directional and strength changes in the dipole field through time, comprise the observed variations in the geomagnetic field that are referred to as secular variation.