Archaeomagnetic dating is based on a comparison of the ancient geomagnetic
field, as recorded by archaeological materials, with a dated record of changes
in the Earth's field over time in a particular geographical area. The
geomagnetic field changes both in direction (declination and inclination) and in
strength (intensity) and archaeomagnetic dating can be based on either changes
in direction or intensity or a combination of the two. Dating by direction
requires the exact position of the archaeological material in relation to the
present geomagnetic field to be recorded, and so material must be undisturbed
and sampled in situ. Dating by intensity does not require in situ samples but is
less precise and experimentally more difficult. The laboratory at Bradford uses
archaeomagnetic dating by direction.
[top]
For an archaeological material to be suitable for dating using magnetic direction, it must contain sufficient magnetised particles and an event must have caused these particles to record the Earth's magnetic field. Many geologically derived materials e.g. soils, sediments, clays, contain sufficient magnetic minerals. There are primarily two types of archaeological event which may result in the Earth's magnetic field at a particular moment being recorded by archaeological materials: heating and deposition in air or water. If materials have been heated to a sufficiently high temperature (>600°C) they may retain a thermoremanent magnetisation (TRM) which reflects the earth's magnetic field at the time of last cooling. Suitable archaeological features would include hearths, kilns and other fired structures. Sediments may acquire a datable detrital remanent magnetisation (DRM) from the alignment of their magnetic grains by the ambient field during deposition. Such an effect allows deposits in wells, ditches, rivers and streams to be dated. However, this aspect of archaeomagnetic dating is still under development, as factors such as bioturbation and diagenesis, can cause post-depositional disturbance of the magnetisation. Archaeomagnetic dating in the UK can be applied to features expected to date from 1000BC to the present day, as this is the period covered by the calibration curve. However, as discussed below the precision of the date obtained will vary according to the period being dated.
Photograph courtesy of Alan Powell.
Samples of robust fired materials are taken by attaching a 25mm diameter flanged plastic reference button to a cleaned, stable area of the feature using a fast setting epoxy resin (Clark et al, 1988). The button is levelled, using a spirit level, and held in place with a small bead of plastecine while the resin sets. The direction of north is then marked on using a magnetic compass, sun compass or gyrotheodolite (depending on site conditions) and the button removed with a small part of the feature attached to it. Samples are trimmed and consolidated in the laboratory with a solution of 10% polyvinylacetate in acetone. Sediments and friable fired materials are sampled by insertion of a 2 cm diameter plastic cylinder, onto which the direction of north is marked. Magnetometers used are sufficiently sensitive for only small samples (c. 1cm3) to be required; approximately 15 samples are needed from each feature and it may be possible to select sampling location to minimise the visual impact, if the feature is to be preserved. Very hard materials can be sampled by drilling cores (eg. Barrett, 2003).
Photograph courtesy of Stephen J. Dockrill.
In the laboratory a spinner magnetometer is used to measure the remanent magnetisation of each sample (Molyneux, 1971). This measurement indicates the relative strength and direction of the magnetic field of the sample. The stability of this magnetisation is then examined by placing the sample in alternating magnetic fields of increasing strength and removing the magnetisation step-by-step. The demagnetisation measurements allow removal of any less stable magnetisations acquired after the firing or deposition event, leaving the magnetisation of archaeological interest. The results of measurements of the direction of magnetisation of a group of samples are represented on a stereographic plot, which shows declination as an angle measured clockwise from north and inclination as a distance from the perimeter.
[top]
Photograph courtesy of Louise Brown.
The magnetic directions from a number of samples expected to have the same date are combined to give a mean direction, the precision of which is defined using Fisherian statistics (Fisher, 1953). (95 represents a 95% probability that the true direction lies within that cone of confidence around the observed mean direction, and would be expected to be less than 5° for dating purposes. A value larger than this indicates that the magnetic directions of the samples are scattered and therefore do not all record the same magnetic field, probably making the samples undatable. The stability of magnetisation of an individual sample on demagnetisation is quantified using the Stability Index (Tarling & Symons, 1967). For a stable magnetisation this value would be expected to be greater than 5, a value less than this would indicate that the recorded magnetisation was not reliable for dating purposes.
Illustration of a stereoplot produced using Fisher statistics the point
marked A is considered an outlier, taken from Fisher et al, 1993: figure 6.20
Once a stable, mean magnetic direction has been obtained this is dated by
comparing it with a calibration curve showing changes in the Earth's field over
time. The calibration curve is compiled from direct measurements of the field,
which extend back to AD1576 in Britain, and from archaeomagnetic measurements
from features dated by other methods. Because the geomagnetic field changes
spatially, data for the calibration curve can only be drawn from within an area
approximately 1000km across and all magnetic directions must be corrected
mathematically to a central location (Noel and Batt, 1990). There is a single
calibration curve for England, Scotland and Wales and directions are corrected
to Meriden (52.43°N, 1.62°W). Conventionally British archaeomagnetic dates are
calibrated by visual comparison to the calibration curve produced by Clark et
al. (1988). However, this method takes no account of the errors in the
calibration curve itself and an alternative method is also used (Batt, 1997,
Zananiri et al., 2007).
The latter method gives a larger error margin on the date but is a better
reflection of the actual error.
[top]
There are a number of factors that will influence the error margins of the dates obtained:
The precision of the calibration curve varies according to the archaeological
period and so the precision of the date obtained will depend on the
archaeological date. As the geomagnetic field has occasionally had the same
direction at two different times, it is also possible to obtain two or more
alternative dates for a single feature. In most cases the archaeological
evidence can be used to select the most likely. Given the number of different
factors it is not possible to give a general figure for the precision of
archaeomagnetic dates but there will be an error margin of at least ±25 years.
It is important to note that, since the method relies on the reliability of
previously dated sites, the calibration curve can be improved as more
measurements become available, so it may be worth sampling a feature even if the
date is well known by another method.
Batt, C.M., 1997. The British archaeomagnetic calibration curve: an
objective treatment. Archaeometry, 39: 153-168.
Clark, A.J., Tarling, D.H. and Noel, M., 1988. Developments in archaeomagnetic
dating in Britain. J. Arch. Sci., 15: 645-667.
Fisher, R.A., 1953. Dispersion on a sphere. Proc. R. Soc. Lond., A217: 295-305.
Fisher, N.I., Lewis, T.L. and Embleton, B.J.J., 1993. Statistical analysis of
spherical data. Cambridge Univeristy Press.
Barrett, J., 2003. Excavations at Quoygrew, Westray, Orkney and the Viking Age
Transitions Project. York University Webpage, http://www.york.ac.uk/depts/arch/quoygrew/
accessed 3.2.03.
Molyneux, L., 1971. A complete result magnetometer for measuring the remanent
magnetisation of rocks. Geophys. J. R. astr. Soc., 24: 429-433.
Noel, M. and Batt, C.M., 1990. A method for correcting geographically separated
remanence directions for the purpose of archaeomagnetic dating. Geophys. J.
Int., 102: 753-756.
Tarling, D.H. and Symons, D.T.A., 1967. A stability index of remanence in
palaeomagnetism. Geophys. J. R. astr. Soc., 12: 443-448.
[top]
Aitken, M.J., 1990. Science-Based Dating in Archaeology. Longman
Archaeological Series, Longman, London and New York.
For general information on
scientific dating methods.
Oldfield, F., 1991. Environmental magnetism- a personal perspective. Quaternary Science Reviews, 10: 73-85. For details of wider applications of magnetic studies
Eighmy, J.L. and Sternberg, R.S., 1990. Archaeomagnetic Dating. The
University of Arizona Press, Tucson.
For an international perspective on
archaeomagnetic dating
Tarling, D.H., 1983. Palaeomagnetism. Chapman and Hall, London & New
York.
For details of the principles and geological applications.
[top]