Home Up




Roman Villa





Magnetomete rsurvey in Sweden

Equipment and Method

We use our own Bartington 601/2 equipment which has two 1 metre long gradiometer tubes spaced 1 metre apart. These each contain two vertical axis fluxgate magnetometers, one at each end of the tube. This enables the top detectors to reject the large scale atmospheric magnetism and isolate the very slight readings which are caused by archaeological features. We chose this equipment as it does not have the dead zones of caesium based magnetometers and can collect data at a faster rate than overhauser based magnetometers.

The survey method normally used is a 30 metre by 30 metre grid and these are surveyed with 1 metre intervals between the lines surveyed and either 4 or 8 readings are taken per metre along the lines. The readings are taken in zig zag mode. Bushes and other obstructions to the survey grids add significantly to the time taken and cost.

Surveys can also be carried out with 0.5 metre spacing between the lines but this takes almost twice as long and the data is often not much better than with the larger line spacing interval.  


Approximately 15  30 by 30 metre grids can be carried out in a day by two or more people depending on the site layout and how many daylight hours there are.  


This is done using TerraSurveyor as some initial processing can be carried out on site and thus influence the direction of further work. This programme is good at removing the half metre stagger which almost everyone gets when using the Bartington logger system, which may be caused by a lag between a reading being taken and when it is logged. The end result is a picture, usually greyscale but colour can be done, and a trace plot of the survey results. The trace plot of original data is often required by curators who wish to ensure that the magnetometers have not drifted (become less well tuned) during the day as some types of equipment are more affected by the heat than others, (fortunately we use one of the more stable sets of magnetometer equipment). The trace plots also help in estimating the nature and depth of anomalies.

Ground Conditions

Magnetometer surveys are dependent upon having soils which have magnetic properties and also having features within them which have different magnetic properties to the surrounding soils.

It is difficult to assess which soils are most likely to show results. Some soils  are not very good but the best way of finding out may initially be to carry out a small survey over a known site and to compare the magnetometer data with that from, say, air photographs or fieldwalking. Even on poorly magnetic soils some indication of previous occupation can be obtained as roof tile and brick can be often detected if they are near the surface.

Sites with metal fences and iron pipelines and other ferrous rubbish can have so much interference as to obscure the archaeology in the magnetometry results  - see the  Requirements page for more detail. It has also been found that previously flooded areas such as the Fenlands can have naturally occurring iron minerals which can obscure the archaeology.


As magnetometry detects very slight variations to the Earths magnetic field, this depends upon how much the ground has been magnetically enhanced and how deep the features are. For most features a depth of approx 1 metre is the best that can reasonably be expected. However deeper features which are very magnetic, such as pottery kilns, can show up at 2 or even 3 metres deep if the ground conditions are right. This method is generally good at finding ditches and kilns and burnt areas but is less good at finding stone buildings. Its advantage is that it is far quicker than resistivity and is less influenced by the moisture content of the soil.